Used to be that when you banned someone for a comment it was possible for the rest of us to read the comment. Now it isn't. Did you decide to change that? I always found it interesting to read them.
IIRC the moderator has the option to choose whether the comment is deleted or if it is merely hidden. Looks like scott decided the former for this one, but not for the one breb links below
Thank you for banning some people, seeing the deleted comment was the reason I checked out the comment section again after mostly ignoring it for a few weeks.
I also asked Claude Code to create me comment fact checking extension for Chrome since the leftover un-sourced lying comments were still annoying.
> So, reportedly the segment was nixed for containing no new information whatsoever and therefore being pushed out in favour of news[...]
Well, a major factor was also apparently that the administration refused to comment.
> since none of this is news and everyone knows about this prison
60 Minutes is apparently the highest-rated broadcast (not cable) TV news program and has been for a long time (although this scandal seems to have hurt them a lot, at least in the short term).
What "everybody knows" is going to seem pretty different to different people. Covering something extensively on 60 Minutes is how your grandma learns about it.
> [...] as in "probably" the Trump admin put pressure on 60 minutes not to air the report [...]
Well, maybe not in the sense of somebody calling them up and saying "don't run this" - as noted above refusing to comment originally provided part of the justification:
But more generally Bari Weiss was installed at CBS by Trump allies and Trump himself has repeatedly praised her because he obviously understands her to be in some sense more favorable to him.
> [...] deporting violent criminals to this prison [...]
16% of the people deported to CECOT had ANY criminal convictions; 4% of the total were convicted of violent crime.
I think this further undercuts your point about what "everybody knows".
Also your position basically suggests that nobody will ever care about the truth again and all news reporting is utterly pointless. At the very least I would suggest the Alex Pretti shooting as a strong counterexample.
Also even if you WERE correct, the Trump administration would have to believe you, but they seem very image-conscious.
> Would you say everyone at CBS and Paramount is loyal to Trump because of this connection?
Clearly not, given the lively internal debate around this story.
> If not, what's special about Bari?
Several things.
Most of her career is a "I'm a liberal, but actually..." shtick where she ends up taking conservative positions and presenting them as either common sense or Things They Don't Want You To Know.
She just got elevated to this position - one for which she's not particularly qualified - by allies of the President.
The President specifically praised her, and he generally has a very strong tendency to praise people on "his side" and denounce people "not on his side", regardless of any other qualifications:
I generally found her explanation for why she delayed airing this story weak and contrived.
> You think there's a reason he had an MS-13 tattoo despite not being a member? Or you think it wasn't an MS-13 tattoo?
The latter, he has some finger tattoos that people claim spell out "MS 13" as though that's an unambiguous interpretation of "marijuana leaf, smiley face, cross, skull" and those aren't just extraordinary common symbols to get tattooed on you.
Gang members routinely have "no criminal convictions" (In America, this generally means they were underage, and haven't done "anything too serious" to get charged as an adult).
Okay. I stand by my general point that it's inaccurate to call people violent criminals when they have no criminal convictions. If the government wants me to consider them violent criminals, they can prove it via the generally-accepted mechanism of trying them for a violent crime in court.
So, instead of just "shaking the jails and sending the ne'er-do-wells home" you're saying we need to do the full megillah? Are you willing to pay for $300,000+ worth of court time? Including appeals?
Please bear in mind "innocent until proven guilty" is a high standard, and that deportation is generally not held to that standard, ever, because we're generally just "sending them home" (and with a nice sendoff song involving marijuana too! La cucaracha).
Trump's gonna say "It's not my fault these folks are being treated so poorly by another country. I run America, not Venezuela or Belize." (perhaps in a larger sense it is, in that he isn't providing them an outlet to escape).
My impression is that new boss Bari Weiss wanted her stamp on it. So it was never "suppressed", it was merely delayed with the intent always being to air it eventually.
Does anyone have coordinates or an address for link #36 about the suburban neighborhood on top of a multi-story building that allegedly exists in Indonesia?
In central Chicago there are now a number of examples of that concept at smaller scale. My wife and I viewed several such "townhomes in the air" in the South Loop area about 12 years ago, and had one of them on our short list, when we'd decided to buy in the neighborhood. (Which we did but it turned out to be a traditional townhouse at ground level.)
Also naive and native (but I think this was a slightly different story — the natural phonetic development vs. learned borrowing from Latin thing happened in French already; English just borrowed both)
I don't think #33 is a good way to build intuition re Russian invasion of Ukraine because
1) It is not true, actual whys have to do with its internal affairs and Putin's regime specifically.
2) It sort of justifies horrendous violation of international norms and the bloodiest European war since 1945. Speaking of which, remember the reasons that Hitler used to justify his aggression, would you recommend them "to build intuition" as well?
I don't think building intuition about why someone would want to do something is equivalent to justifying it. I think a map of German population in the Sudetenland would successfully help explain why Hitler wanted to annex it and be a useful historical contribution, without suggesting that wars of aggression are okay.
But isn't it clear that German population in the Sudetenland was merely a pretence? And when you say that this pretence is "building intuition about why someone would want to do something" you are indirectly white-washing them, because you invite us to think of them as a reasonable person that has reasonable arguments instead of a murderous psychopath that makes these arguments on the fly.
I don't think the pretense-real-reason distinction is as strong as you think. I think you can be an aggressive warmonger, and your choices of which aggressive war to pursue will be based on various ideas about national glory and revanchism and how to get your way, and those will depend on facts of history and demography.
Putin claimed that a meme posted on an American Website that just happened to be based in the Ukraine was part of the justification for invading the Ukraine.
Either that's some really good pretense, or Russian intelligence was really off its game.
Did you hear how America/Western Europe spent 10 years wetting their pants about the USSR invading past the Iron Curtain*, when the USSR couldn't do more than drive a few rusty tanks around (and was doing so, in order to attempt to intimidate a West that they very well knew they couldn't fight?)
*absolutely certain it was coming "any day now"
It seems like America cannot get a good read on Russia. This last war (Ukraine, currently still limping along fighting) America thought Russia was strong militarily, but very weak economically (so they thought sanctions would crush Russia in 2 months. Years later... Russia's held strong).
I agree, the casual chain from perceived loss to revanchism, to fascism, and then to external aggression has some explanatory power, both for Hitler's Germany and for Putin's Russia. Although I think that it is important not to skip the intermediate steps in the explanation and to make distinction between reasons as social forces, dictator's personal reasons, and reasons that they present to us as justification.
Also don't forget that people tend to believe their own propaganda eventually. It is hard to live with cognitive dissonance. So at that some point both Hitler and Putin thought that liberal democracies are basically a bunch of sissies who don't have a real man's backbone.
One should always start with the idea that someone is a reasonable person with reasonable arguments (like, say: "mandatory gym class" to cite one policy proposal from Mein Kampf).
Only when they provably make very stupid decisions, like "Protect the coke factory at all costs!"* (despite it being of next-to-zero strategic importance, as it doesn't make end products, and the factories that do make the end products are destroyed), ought one to consider whether or not they actually do understand that coke doesn't always mean cocaine.
He had other goals but that doesn’t mean that it was a “pretense”. Everything we know about Hitler suggests he was strongly motivated by the idea of a united German people. There’s no reason to think that this wasn’t the case.
He wasn't a "white nationalist", every European country was white, he was a GERMAN nationalist. Slavs (such as Poles) are definitely white, but they weren't included in his nation.
He had multiple reasons for invading. For example, Norway got invaded because Hitler (rightly) expected the UK to violate its neutrality and take advantage of its coastline, Hitler jumped the gun and grabbed it first so he could protect ore shipments from Sweden (which he didn't have to invade).
Poland was next, and Poland was also a country with a big German-speaking area.
And then all subsequent invasions starting with Denmark came after WW2 had kicked off in earnest and were intended to outflank or defeat France and Britain.
(Again, acknowledging that these things happened for some actual reason doesn't make them okay.)
Poland did, as for France, they declared war on Germany, not the other way around, Greece and Yugoslavia was Mussolini's stupidity, so the only ones that really count are Denmark, Norway and BeNeLux, now those are truly hard to understand. I think Hitler did not simply want Germans in one Reich, but rather "Aryans" - this is clear from the table talk book, that nationalism was really just public propaganda spread to placate the conservatives, and the real goal was always racism.
The German population in Bohemia and Moravia was quite significant, had lived there for 1000 years, and under Habsburg rule had generally happily considered themselves ethnically "German". After Austria-Hungary collapsed, the Germans in Sudetenland assumed that of course they would be joined up with the other Habsburg Germans in Upper Austria, Tyrol, Carinthia, etc. in a state called "Deutschösterreich" ("German Austria") and were quite shocked and angry when that turned out not to be the case. Hitler grew up in an Upper Austrian town right across the border from German majority towns in Bohemia, and seems to have had Bohemian ancestry. So no, I think this was a cause that was actually very dear to his heart and that incorporating Germans that had been "unjustly" cut off from the homeland by Wilson's arbitrary borders was sincerely one of Hitler's main goals. Incorporating Sudeten Germans into the Reich wasn't a pretence, it was one of the issues that radicalized him in the first place.
Almost everybody is a reasonable person in their own version of the story. I really recommend reading Gaetano Mosca here. Basically, feeling like a bad guy is psychologically unbearable, everybody comes up with a story that makes them look like the victim or the hero, even someone like Hitler or Putin. There were genuinely good people who were still faithful Nazis: https://en.wikipedia.org/wiki/John_Rabe I know this is hard to imagine, but these were the facts.
Respectfully, I think you are wrong here and wrong to signal boost that info graphic.
This map and the Sudetenland example build sympathy, not strategic understanding. The target outcome for the audience is "if I were in their shoes, I would do the same, thus it isn't so bad."
To further see the distinction, would the ethnic German fraction of the population explain other WW2-era German invasions: Poland, Benelux, France, etc? No. This was not an explanatory factor with predictive value. On the contrary, we know at the time it served as cover to a (false) hope that the Sudetenland aggression was a one-off.
With respect to the USSR->Russian Federation info graphic, the framing plays on the intuition that a something (country, capital city?) has a right to (control, oversight, governance?) of "its" citizens and economic production and that the post-USSR dissolution was a loss in roughly the same way that an individual losing property is a loss. I understand that the grey text explains and partially justifies, but that visual choice is pretty telling. The explanation text is extremely hard to read and is not what most of the audience will take away.
I agree. A better comparison would any other past empire wanting to invade countries that have since gained independence. E.g. the UK invading India, Kenya, or the US.
The UK doesn't border such regions. You could discuss the time period when English kings held part of Normandy and fought France for territory for a long time (the "hundred years war" for example).
Why is that significant? The original comparison ignored things such as culture, language, and how long these nations have been a part of the empire, so we might as well ignore proximity too.
I said taking into account culture was a good point https://www.astralcodexten.com/p/links-for-february-2026/comment/210410874 And sharing a border with a country obligates some kind of relationship (sometimes friendly, sometimes hostile, sometimes a mix) in a way that being an ocean away doesn't.
Northern Ireland does indeed border the Republic of Ireland, which is why the UK is more entangled with them. Ireland was not in the list of "India, Kenya, or the US".
I am definitely not a Chamberlain expert, but with respect to the PM himself, it seems there is a big range of reasonable interpretation. I endorse the view that Chamberlain was not a one-factor thinker who was overwhelmed by the idea that "he is only interested in ethnic Germans and will stop at the Sudetenland."
To me, however, "went straight home and started rearming" is misleading. It sounds like strong evidence that Chamberlain gave no meaningful weight to that argument, but I don't think it is supported by the actual defence spending stats.
The rearmament defense spending had been, arguably, underway for several years prior to the Munich Agreement. Data I found, link below, shows the following UK defence spending from 1934-39:
1934 120 mm
1935 120 mm
1936 150 mm
1937 200 mm
1938 210 mm
1939 270 mm
(the data is presented in GBP bn, which I am interpreting here in millions, but I hope this doesn't give the wrong impression about the level of precision of the data)
The increases from 1935 to 1937 were larger in absolute and relative amounts than in 1937 to 1938. Also, it might be useful to see these in context of the defence spending that suggest the 1938 budget of 210mm was not a particularly hawkish stance:
1920 690 mm
1921 300 mm
1922 220 mm
Qualitatively, the public addresses (like the Jan 1939 Birmingham speech) suggest that Chamberlain did have significant hope that German expansion would stop at the Sudetenland.
The ethnic German population in Poland in 1939 was significant enough that, yes, you can explain that invasion as well based on ethnic reasons. Danzig was a majority German city, Western Poland had been under German rule until 1918. Southern Poland had been under Austrian rule and there were still plenty of Germans in Polish Silesia. Kattowitz was German enough that serious fighting took place when it was assigned to Poland by the allies.
I will go farther and say that the "they used to control it, so they are likely to invade it" argument does work for Poland (I am not sure if it works for all of the territory they took in 1939 or not.)
Between Sudetenland and Poland, though, was further occupation of Czechoslovak territory. That seems to falsify the claim that Germany was *just* restoring former territory and/or areas with high ethnic German concentrations. No?
The annexation of the Protectorate of Bohemia is indeed why Chamberlain decided Hitler wouldn't abide by any deal, so the next step (Poland) meant war.
I think that was a mistake. There is nothing specific to Russia here. China claims Taiwan for the same reason. They also claim (unofficially) the borders of the Xing dynasty. It's perfectly normal for countries to claim past territories and territories with similar ethnicity. The US wasn't particularly thrilled to lose the south when it seceded either.
I was always irritated, then disgusted by Russian chauvinism (lived and worked in RF/UKR for 9 years between1994-2008). I would give my life if I could take Putin's. But, sure, an interesting map (not a territory!); thanks for sharing!
I don't "sympathize" with Putin, I think he's awful. The map conveys info for Americans who lack intuition about the size of such countries/regions within the former USSR.
The idea itself come from this https://digital.library.cornell.edu/catalog/ss:19343225 which was a direct propaganda campaign against the Treaty of Trianon. I’d be 90% sure whomever made the map did so as pro-Russian-expansionism propaganda. That doesn’t mean it’s false, bad or unworthy of existence (Casablanca was propaganda) but I’d be cautious about passing it on uncritically.
Notably as well, they’ve done it in such a way as to make ugly janky borders, as opposed to modern Russia’s basically cromulent boundaries.
That map acts as if culture is meaningless. You can't compare New England to the Baltics, New England is arguably the cultural heart of America. Losing New England and Boston would be more like Moscow losing control of St Petersburg and Novgorod. A better comparison to the Baltics is probably Hawaii.
I will accept Belarus = Ohio because most Russians don't understand intuitively why Belarus is an independent state. Ukraine is trickier. Kiev is in some way the heart of "Russian" culture, but it is has never really been integrated into Muscovy, which is what the modern Russian state actually is. Texas is probably a much better comparison than California - the Lone Star state has a historical memory of not being part of the US, Texans have a strong local patriotism and a lot of Texans feel about the federal government the way Ukrainians feel about Moscow. Also Texas has played an outsized role in the way foreigners see the US, just as a lot of Ukrainian cultural baggage (cossacks, borshch) is associated by foreigners with Russia.
Almost no one in Russia has any nostalgia for bringing back the Stans or the Caucauses. I suppose you could posit that a rump US based around the historical 13 colonies and midwest might feel that way about Lousiana and Mississippi but really Puerto Rico and Guam are better cultural comparisons, even if economically less significant. There aren't even significant Russian minorities left in those places other than Kazakhstan. Most of the Russians in Georgia are recent arrivals.
Britain had a uniquely large accent diversity, due to the extreme balkanization of small towns in Britain. France, in comparison, had a much less extreme version of "accent diversity" because towns regularly knocked knees with the next town over.
I've never heard that before, why would British people have been so much less mobile than French?
Britain also has extremely strong class-based accent differences.
I don't know enough about enough countries to say what's normal and what's not, but I will note that Germany also has strong regional accents, and I think the vocabulary has more regional variation than in English as well.
Paranoia (this is the standard reason given for why English cuckoldry was often done within the village, as opposed to the French standard of "do it the next village over, as an affair")? Lack of long-distance rivers, which made travel much less profitable? England's civil wars might have also had something to do with this (France was more established as a "real country" for a long time before England actually managed to weld itself together).
"Class-based accents" are a later addition, coming from when Received Pronunciation was developed (unless you're speaking of Norman versus Viking accents, which was a whole different ball of wax).
> I've never heard that before, why would British people have been so much less mobile than French?
They wouldn't.
France, like pretty much everywhere else, made strong efforts to centralize the language spoken within its borders, which was significantly 𝗺𝗼𝗿𝗲 diverse than what we see in England. English in England is also much less diverse now than it used to be in the past, but it never achieved what the Romance languages did, because England is a tiny place.
As TGGP originally commented, there is more diversity of English in England than there is in the USA because English has a longer history in England.
It is said that in the alps, people in one valley couldn't understand the people in the next valley over, for both German and French dialects.
However, unlike English, German and French both have sort of central authorities defining what the language is, which I think probably contributed to the reduction of these differences.
Russia has hardly ANY accent diversity - including Russian speaking parts of Ukraine, where I lived some years (only one I heard of was: Is it ARboos or arBOOS - for melon), while I can not understand southern-US talk without subtitles. Just saying. Also, ofc: PTN PNX (ie: screw Putin)
This is a very good comment, and Texas == Ukraine gets to a lot of the heart of "Why Russians think Ukraine should be theirs" and why the Ukrainians do not (but might allow that Russia should be theirs, and controlled from Kiev).
(Puerto Rico has a lot of cross-cultural stuff with New York City, but that's mostly Puerto Rican immigrants, not the other way round).
>Almost no one in Russia has any nostalgia for bringing back the Stans or the Caucauses
Chechnya, Dagestan, and a few other obscure Caucasus republics are part of Russia now? Famously, Russia fought two very bloody wars to keep them. Did you mean Armenia & Azerbaijan instead?
The parts of the Caucases that lie outside the Russian Federation. You could compare Russian feeling towards Georgia to the way U.S. feels about Mexico. It’s definitely not America but we won’t tolerate it being allied with an enemy power, we like to vacation there and the food is not just a lot better, it has also become part of our culture. New Mexico is like Dagestan. It was Mexico, has a different culture but we aren’t letting it go. Sochi is like LA. It’s been firmly integrated and is now just Russia, the way Americans never think about how a town with a Spanish name is just part of the U.S. now.
#33 could have done a better job of matching the approximate cultural as well as demographic and economic relationships, but it does I think a pretty good job of showing how Vladimir Putin is basically channeling Londo Mollari at his worst (https://youtu.be/YbckvO7VYxk).
And it's an explanation, not a justifcation. Putin remembers Russia at the height of its imperial power and glory. He was there. He was part of it, and it defined him. Now it's gone, so no, he's not going to care whether it's "justified", he just wants it all back the way that it was.
The map, translates "the way that it was", into terms Americans can easily understand.
This is a poor characterization of Putin, and of Russia's stance in general. They are not pulling "Fortress Russia" out, as their playbook. (America on the other hand...)
"Above all, we should acknowledge that the collapse of the Soviet Union was a major geopolitical disaster of the century. As for the Russian nation, it became a genuine drama. Tens of millions of our co-citizens and co-patriots found themselves outside Russian territory. Moreover, the epidemic of disintegration infected Russia itself."
I also don't think it's a good way to build intuition, but not for the reasons you stated. I just don't get what intuition it's supposed to build. How much weaker Russia is compared to the USSR? But everybody knows the USSR was a superpower during the Cold War and Russia was barely a great power after the collapse. The relative size of Russia and the former Soviet republics? But I knew that already; there's a reason Russia is still a lot more powerful than Ukraine or Latvia.
I do not buy control systems as the reason for the Boeing Jetfoil's failure, because it is clearly and transparently false. Yes, in the mid-70s, computers were rare and expensive by consumer standards, but this was far less true in industrial applications. The F-16, which was fly-by-wire (similar technology for planes) first flew around this time, and Boeing kept delivering Jetfoils up through the 90s, when the program was sold to Kawasaki (because it wasn't profitable enough and most of the customers were in Japan.) They delivered more, which remain in use, including one delivered as recently as 2020, and another one is currently on order. If it was all "control is too hard", they could easily have replaced the system at some point in the last 30 years. They probably have. Also, not mentioning that they're still around or the reasonably long history of hovercraft ferries in other contexts makes me very suspicious of the startup's story. (Although this is a startup, so it's not like I'm surprised.)
The other thing that doesn't get mentioned is high maintenance costs. Yes, not needing engines will help some with that (and even engined hydrofoils will get better because they won't have to do silly things like pump water up into the hull where the engine can push on it, then back down to the outlet) but there's also stuff like foil erosion that makes them really expensive to run. Maybe they're going slower enough that's less of a problem, but I'd want to see some longevity tests first. The 6-passenger market is definitely a better spot for them than anything big (I suspect one of the big killers in the ferry market was that it can't take cars) but it will be interesting to see what comes of this.
Edit: As for naval use, I'm very skeptical. As always, my question is "OK, what systems exactly are you mounting on this, and what do they do?" Maybe there's a niche for inshore use, but I really don't expect a comeback for the military hydrofoil any time soon. There are options with much better track records that we aren't using.
Setting aside whatever issues there may be with the technology, I just don't get the economics of this. Apparently the largest version seats nine passengers. There are obviously cities where ferry service can be a fairly significant portion of public transportation, but the ferries are much, much bigger. If you want to run a whole lot of ferries, you eventually run into traffic issues, especially in terms of dock space. And you need good connections to other forms of mass transit at both ends (or you need to have docks that are within walking distance to major destinations). I guess there are niches where this could be useful, but most of the niches would probably be popular enough to be served by a bigger ferry.
Based on the article, they're talking about Uber/Waymo for water taxis. Which might not be completely crazy in the Bay Area if you can charge a premium for Uber-hydrofoil-Uber over just doing Uber directly, particularly at rush hour. But yeah, there's going to be scaling concerns, and while the obvious solution there is to make a bigger hydrofoil, we've sort of had those for a long time...
Navier's actual marketing seems to be to people who want a hydrofoil boat.
I guess it works if you can guarantee that the wait times are minimal for every leg of the trip. If you have to wait ten minutes for the second and third legs of the trip, you're probably eating into a lot of the time savings. And this boat is going to cost a lot more than a car, so you need to charge a much higher price for an equivalent return on investment.
There are a few cases where they work well (Staten Island and Hong Kong, as well as the river in Brisbane) but they have an inherent limitation in that every station has no development possibilities on one side (being at the water) and the ferry terminal itself needs to be at some of the most expensive land in the city (waterfront).
Another issue is that many of the places where all these criteria are met are busy waterways which have modest speed limits anyway, so the hydrofoil is useless.
I agree about the bottleneck being something other than the control systems, but I do not think the high maintenance costs are unsurmountable either. For example, Russia is using a large fleet of Raketa/Meteor/Kometa hydrofoils as water buses on its lakes and rivers since late 1960s. They are passenger only, but seat between 60 and 120 passengers, so are not quite at the taxi scale.
I am well aware of those, having done a deep dive into the uses of the hydrofoil about 6 months ago, but I'm not sure they prove your point. Those were produced in the Soviet Union, and cost-effectiveness was not always high on their list of priorities. Now, on the other hand, they did have some export success, but I don't remember how long those tended to stay in service after being sold abroad. And since the 90s, it could easily be "well, this is what we have, keep it working".
The other issue is that I don't really understand the foil erosion problem that well. It was definitely an issue for the Pegasus class, which I have by far the best information on. But how can it be mitigated? Was it mostly because the Pegasi were so fast, and cutting speed makes it go away? Was it because the Pegasus was submerged-foil and the control system couldn't handle the loss of lift from the erosion? Or just that it had no margin to handle the loss of lift when trying to get up on the foils? A smaller surface-piercing foil at lower speed addresses all of these, but I'm not sure that the real answer is even on my list.
Modernised versions are still produced and used in Russia today, e.g. Meteor-120R, launched in 2022. Cost effectiveness was not much of a concern in Soviet times, but it is as much a concern to current Russian operators as it is to anybody else. In Estonia and Lithuania, Raketa produced in 1960-1970s stayed in service until early 2000s (https://en.wikipedia.org/wiki/Raketa_(hydrofoil))
As far as I can tell, it's just Meteor-120R, and that seems in very limited production. And even there, I could easily believe that in a few cases, it's cheaper to build a new hydrofoil ferry than to redo a city's transportation infrastructure to use a different method. Which is not the same as saying it would have made sense from the start. But this whole thing is deeply mysterious to me, and will remain so unless someone runs across a careful academic analysis of the operating costs of various high-speed ferry options. I'm not looking, because that series is done (ish) and I have other things to write about.
57. I use Threads. For some reason Threads became popular as a Twitter alternative in Finland (perhaps it was simply launched here at the same time as there was one of the Twitter exoduses?) and the community is probably currently better there than on Twitter, so I've moved there rather organically as well. There are probably some other nations where this has happened as well, just like Twitter communities of some other nations have moved en masse to BlueSky, Mastodon etc.
The Finnish Threads community is basically stereotypized as "horny late millennial white collar workers", ie. what Twitter used to have before the... acquisition. In my experience the American Threads community seems to be heavily African-American, discussions on topics related to this community pop up on my timeline all the time.
My first thought reading this was "that's an insane thing to not know for an educated adult American" but then I remembered that you're a West Coaster who doesn't browse the Caribbean for quick vacation getaways and also that I live in NYC, with a very large Caribbean population.
On #57, I think it is just surprisingly easy to be unaware of the social networks "other" people use. If you use Twitter it is probably because most of your network uses it, and things like that tend to be sticky. Personally, I almost never use Twitter (and have been that way since way before Musk took over), know very few people who talk about using it, and have no real awareness of how many people actually do use it.
Regarding murders and trauma care, what percentage of murdered people used to die in (or on the way to) a hospital?
Improved trauma care only helps if you receive trauma care. If back in 1980 only 10% of murdered people died after an ambulance arrived then care could become instantly life saving and still barely impact the murder rate.
Seems like a potentially debate resolving question.
I had the same thought. In some cases, "murder" is essentially an assault attempt whose consequences exceeded the perpetrator's intentions. But more often, I think, murder is the result of the perpetrator deliberately killing their victim, and they'll stop when they feel adequately assured that their victim is dead.
Some types of assault, like knife attacks, are more survivable than many people suppose, and you can survive being stabbed once, or several times, with high likelihood, but that was already true decades ago, and people committing deliberate knife murder will usually frantically stab their target over and over until they stop resisting.
I don't think most murders are "intentional outcome" murders. I think most murders are part of dispute resolution, in cases where cops/law enforcement cannot be called. And there, folks ain't trying to kill the other guy (as that gets his gang madder'n hornets)... You get "heat of the moment" reactions, and that leads to "calling 911", for damage control if nothing else.
Do you have a source on this? I've never been in a gang, but I've lived in high murder rate/ high gang activity areas, and talked to people who've been in them (in some cases,. maybe still are?) And this doesn't correspond to the experiences they've related, at least.
Not that sub-murder gang violence doesn't occur, but I've heard a fair number of firsthand accounts of violence, and all (except possibly one uncertain case) which led to death were intended to.
Sources yes. There was an incidence of "gangland violence" where "bring back a bloody knife" was the ask, in order to join the gang. Apparently, ganging up on the local dingus* turned out to be the actual "what happened." He was hospitalized, as there were more than one kid who wanted to join the gang.
You have "beefs" that get "resolved" at neutral places, like bars. You'll have "violence resolution specialists" that provide counseling on "how not to draw guns when drunk" and other things like that. This wouldn't be a "big thing" if there weren't deaths/bad injuries to prevent (https://scholarworks.lib.csusb.edu/etd-project/3209/)
Yes, there are actual "murder this guy" jobs, and people do have actual jobs as assassins as well, sometimes under the employ of state actors. Those are relatively rare, compared to "everyone who has a job in the gang."
There are a few other things going on -- you may not be getting the whole picture, on "trying to kill someone" (there's incentives to say "yeah, I totally meant to kill him" as it makes you sound badass). Said critique applies to my sources as well, "bleeding hearts" may do interventions that save only a small fraction of kids, simply because "one kid saved" is a net good.
And I've discussed with a source about LA gangs calling 911 (we're not talking about ethnic cleansing, which would obviously not involve calling 911, as the objective is murder).
*when a little guy asks the drunk, "are you afraid of me" we aren't exactly surprised when he says "no" -- but when the little guy draws a knife, the correct response is not to continue insisting no, but to RUN.
This is all largely familiar to me, but also all seems consistent with "most assault attempts and instances of violence do not lead to death, most deliberate murder attempts succeed (or at least are not prevented via quality of medical care.)"
Your account of gang life is fascinating and I want to thank you for sharing it, but aren't most murders committed in the context of a domestic violence situation? In which case violent crimes and assaults could be going down because as a society we have learned how to handle domestic conflict more effectively.
I'm not sure. You have business, and then you have "reputational business" (which could be DV if you stretch it -- shoot the woman at the bar, because she's disrespecting you by flirting with another man -- or shoot the man, because he ought to know better too).
I'm not sure in a country with a 3x suicide to murder rate that we know where most of the dead bodies are. And that's not getting into the "excess Deaths."
Don't fall for the frame until you've investigated all the "easy avenues" to call something "not a murder."
While reliable statistics are hard to come by, as is often the case with anything involving guns,* it appears that about 80% of shooting victims survive. That’s for handguns; rifles and shotguns have much lower survival rates, but are seldom used in crime. I do not know whether the 80% figure includes self-inflicted shootings.
* = Undoubtedly for political reasons no agency keeps records on how many justifiable defensive shootings by non-police occur each year.
Another study showed gunshot wound mortality going from 16% to 10% among admitted patients over the course of 20 years (https://pubmed.ncbi.nlm.nih.gov/28697020).
Back of the envelope math: 30% do not get an emergency response. 41% get a response but die before reaching the hospital (70% x 16/27). 29% get a response and die at the hospital (70% x 11/27).
Dropping mortality from 16% to 10% on admitted patients only reduces deaths for the 29% that arrive alive. That would be an 11% overall reduction in murders from improved care.
Take away: I put my estimate on murder reduction due to increased trauma care at 20%, with a range of 10-25%. My data was from different places at different times but is consistent with other articles that I read over in my admittedly quick review.
A simple extrapolation of 11% over 20 years is 27.5% over 50 years, pretty close to the claim that "murder and other violent crimes have “only” gone down by a third in the past fifty years".
Linearly extrapolating the gunshot mortality rate going from 16% to 10% is probably wrong, especially if you're extrapolating that far (30 years). If we extrapolate 34 years past the end of the study, the mortality rate will be negative.
EMS provide some emergency care on site and while in transit and I would expect the outcomes of this care to improve alongside hospital outcomes improvement. So just counting the 16% to 10% improvement in post-admission outcomes underestimates the overall improvement of victim mortality from medical advances.
Of all homicides, 70% had an EMS response. Of all EMS responses to gunshot/stabbing victims 27% died (16% before hospital, 11% at hospital). Because we are only looking at deaths, that 27% death rate represents the entirety of the 70% with of murder victims who had an emergency response. I applied the ratio of where people died after an EMS response (16/27th before hospital, 11/27th at hospital) to the 70% of murder victims that had an EMS response.
Robbery-murders seem way less common today than in, say, 1980. I seldom hear about 7-11 clerks or mugging victims in Central Park being shot by armed robbers compared to a couple of generations ago. Armed robbery seems down in general, perhaps because there's less cash payments compared to electronic payments.
Our society is better set up to discourage rational crime than it was 60 years ago when crime was just beginning to rise, so most murders these days are due to stupid people getting in stupid beefs with each other.
Trauma care won't help with intentional murder - if the goal is killing, the perpetrator will indeed shoot the corpse. It should reduce manslaughter incidence - especially vehicular manslaughter. I have no knowledge of what technical crimes are included in "murder rate down" statistics.
Vehicular manslaughter at speed tends to lead to people with enough trauma that EMTs say "died instantly." This is nearly always a lie, but did you really want to tell the grieving widow that her husband died screaming as he burned alive?
On the murder survival rate, thing; the increase in survival for assault victims is real... but it mostly happened decades ago (70s, really); there has been little change in the last 40 years or so. You'll often see people quoting assault survival rates that exclude people found dead before reaching medical care, which rather muddies the waters.
Recent trends have actually been for *worse* trauma survival rates, not better.
As a fellow pedant, I stopped correcting people on the champing/chomping thing when I realized the latter is just an American regional variant of the former. I’d never correct someone for saying “I don’t give a rat’s arse”, so why bother with the bit chomping 🤷♂️
Even absent the regional meaning, "chomping" is a pretty good description of a horse biting a metal bit! It's an intelligible phrase in its own right and has a nearly identical meaning, like saying "throw the babe out with the bathwater" instead of baby, or "put him on a plinth" instead of pedestal.
It's not good style, obviously, but it shows a familiarity with language and metaphor that's absent in phrases like "could care less". It's arguably more consistent with the rest of the language, too.
Interesting that I saw how coding LLM did 1+1 calculation without any reason during thinking. It was in JetBrains IDE chat interface. I am puzzled because this was not a tool, it was just directly during reasoning. But it might be related to the calculator tool example somehow.
It's probably not serious anyway, but your dating startup would instantly fail due to nearly all men giving a 9 to nearly all women, and vice versa nearly all women would give 0 to nearly all men. See swiping rates per gender.
I would imagine it would scare away most people because rating from 0 to 9 is too complicated, and those who remained would use it more as intended.
Related: OKCupid, before it became a Tinder clone with fake accounts, was actually a very good dating site because it scared away people not willing to write a good profile and answer questions (at least for me; I got more dates out of it than from any other site, an most of them were at least interesting).
If users aren't encouraged to write good profiles or include much in-depth information about themselves, I wouldn't expect the 0-9 rating to be very meaningful as a predictor of compatibility anyway.
Correct, though what's the best rationing system here? Intuitively one may ration by high scores, but arguably, both very low and very high scores should be rationed.
> Correct, though what's the best rationing system here? Intuitively one may ration by high scores, but arguably, both very low and very high scores should be rationed.
I think on the dating app side, it's already pretty good - there's apps like Raya and The League that have entry barriers (at least on the male side, I think on the female side you just have to be hot, basically), and sorting by quality at the app stage is basically what you need for a stable matching problem to work.
The big problem is that the non-selected dating apps don't work for below top-quintile men and women to actually pair up in LTR's.
And for that we need something else entirely - we need matchmaking.
All the matchmaking things in the US kind of suck - tiny user populations, high costs (certainly higher than 80% of people will pay), and so on.
So what we really need is a cheaper, high quality matchmaking thing that can apply to the 80%. Personally, I'm hanging my hat on AI.
When everyone has AI personal assistants in their ears that know everything about them, it should be pretty easy for the AI's to help match people up, in a way that's much higher fidelity and better compatibility than the people themselves manage.
I was thinking more in terms of Scott's specific proposed dating app, less on the general dating app market. Otherwise I have little to add here, since I have almost no experience with contemporary dating apps and no inclination to change that.
On the AI personal assistants, I increasingly fear that the same people currently struggling to find a partner will just forgo finding the irl romantic partner and just pretend that the AI personal assistant is one.
> On the AI personal assistants, I increasingly fear that the same people currently struggling to find a partner will just forgo finding the irl romantic partner and just pretend that the AI personal assistant is one.
Yeah, but to be honest, that sounds like winning to me, too?
Literally every generation of women since the 50's has had lower marriage rates:
This is largely a good thing, because the vast majority of people suck. When ~90% of women were married in the 50's, it was an immense charnel house of suffering and wasted potential.
How do we know this? Even *today* divorce base rates are ~42% for first marriages, and about half of the remaining marriages are net miserable for at least one party. That's a roughly 2/3 failure + misery rate. Think how net miserable those 90% of marriages were, that could NOT get divorced, back in the 50's!
As soon as no fault divorce opened up in the late 60's, divorces surged and that was a good thing - it was millions of soul-crushing relationships suddenly able to be dissolved! Not to mention that all the actual female talent that had been shackled to soul-crushing duds were now free to get careers and open bank accounts and live more meaningful lives, which was also great for GDP.
And likewise, the fact that each subsequent generation of women has married at lower rates is a great thing! That's millions of shitty relationships, not being entered into! It is very much a "that which should be destroyed by the truth, should be" sort of scenario, and we should rejoice that it's happening.
So when this trend is accelerated by AI, who will be superhuman at relationships in basically every respect (at least when whatever soon-to-be trillion dollar company gets good sexbots worked out), that TOO will be a good thing, for all the same reasons.
Fertility crisis? We'll have ~8B people through year 2100. Serious declines don't hit for another 100 years. 200 years is more than enough for AI technorapture or transgenic pigs or some other artificial womb solution to get worked out.
Humans aren't fungible. Here in germany, the age pyramid has already gotten so lopsided that it's entirely impossible to actually realize the retirement promises that the state has given to the public all their life, and this will most likely get worse in the 2030s. Immigration exacerbated this issue, instead of improving it as hoped. In the longterm evolution will sort things out, as always, so I'm not particularly worried, it's the near- to medium-term that looks bad. Depends on the country though, the US seems relatively fine.
>Divorces ...
I guess I just simply do not buy into the entire framework. I grew up in a conservative christian region, with marriage & divorce rates more similar to the 50s then anything else (though the legal environment was pretty much modern already). I became an atheist and moved to a liberal university, where most of my new friends unsurprisingly came from a liberal background, and now that I'm older I also get to see urban liberal coupling patterns up close.
And the more I see, the worse it looks. The higher percentage of people who break up aren't happier; They just falsely blame their problems on their partner, falsely believed that a single life and career/self-actualisation makes them happier or falsely thought they could "do better". I'm aware of surveys on the topic, but I have absolutely zero confidence in their reliability; Nobody has seemed more miserable to me than a childless single women in her 40s proclaiming how much she enjoys her freedom. And they do occasionally admit that they regret their decisions, though rarely publicly, and rarely for long.
The kids also suffer a lot under a divorce, and making family life nice simply isn't just about the right match (though it certainly helps); It's about adapting to your partner/kids and putting in the work. About priorities and just letting go of your individuality (not entirely, of course, but partially). As far as I can judge it, liberal reddit-tier "always focus on your needs and break up over every inconvenience, you go grrrl pwr" is the worst thing that has happened to human happiness in a long while. And this even partially applies to liberals who do stay together, because they regardless just can't help but always blame their partner, always dream about how much better their life could be without any obligations or with someone "better". And it's also part of a bigger package that means well but just makes everyone unhappy, such as therapy and guilt culture.
The conservatives aren't perfect, of course, and their divorce rate isn't zero, either (which is good - I'm not against divorce in general, I just think liberal culture is far too biased in favour of always breaking up). But even my wife, who comes from a more liberal background and had lots of ultra-liberal friends in her early adulthood, just can't shut up about how nice, happy and well-adjusted everyone is when she meets my family or my old school friends. I know lots of liberals are really into the idea that it's all some stepford smiler hellscape, and some people like that surely exist, but no, you actually can just be happy with the people you have instead of making yourself and others miserable by always trying to get something better.
But I'm aware this is something that probably can't convince anyone who hasn't personally experienced the difference.
Well... sort of. One of the mutually-incompatible criteria is monotonicity: ranking a candidate higher in your personal order of preference should have a nonnegative effect on the electoral outcome of that candidate.
But this isn't a problem for the system you're talking about, where you rate people on a scale from 0 to 9. Arrow's Impossibility Theorem doesn't apply to that at all, because you're rating people in cardinal numbers.
He’s mentioned in the past that this is designed for people you know in real life, and aren’t passionate about but would be sort of interested in. I don’t think normal swiping rates per gender would apply to real acquaintances.
That's a whole 'nother can of worms - I remember an app that would allow you to "secretly" check a box for facebook friends you'd be willing to date, and they would only see this if they *also* get the app and *also* check that box. Seems simple and uncontroversial, right?
Except when some women heard about the existence of it, they'd get the app, check everyone's box, and then post publicly about those disgusting creeps who checked hers. Dating among irl acquaintances really needs plausible deniability and/or very well-honed social instincts, especially in modern culture.
> Except when some women heard about the existence of it, they'd get the app, check everyone's box, and then post publicly about those disgusting creeps who checked hers. Dating among irl acquaintances really needs plausible deniability and/or very well-honed social instincts, especially in modern culture.
Did this really happen? Seems obvious that some will exploit the system in that way to gather information, and sadly probably inevitable that some will make it public. But did any actually frame it as "disgusting creep" rather than (also bad, but at least not moralised) "lol, loser"?
I can see the chat, but when I got to the question about terminal lucidity the response was only half generated. Scott, did you get a full answer that's just not coming up for me?
When I try to view the linked article "Barsoom - Amelia Sans Merci", about a British government-sponsored project to discourage people from researching right-wing ideas online ... I get a message telling me this is "Age-restricted content" and I must verify my age to continue. Substack helpfully explains that this is required to comply with British government regulation (the "Online Safety Act").
In other words, in order to view subversive content, I have to identify myself. This seems like it might cause a chilling effect.
In the case of auto-accident insurance, there's a perverse incentive for claimed damages to adjust up to or just below the defendant's insurance limit, because the defendant will probably just settle at that level if it makes everything go away (modulo higher insurance rates going forward), but will fight if the alternative is a big chunk of their life's savings.
Dialing your personal liability insurance up to $1E6 rather than $1E5, has a very small possibility of protecting your assets in the case where you e.g. really did inflict $1E6 of medical bills on someone and they have the receipts and really need the money. But in the much more likely case where you are involved in an accident and the other guy's medical bills are say $20,000, the answer to the question "now how much 'pain and suffering' should I ask for on top of that?" will tend to be $980,000 rather than $80,000.
Demanding that *other* people carry $1E6 in liability insurance, looks a lot like a gift to plaintiffs and their lawyers.
I can't think of an answer that doesn't involve large-scale tort reform.
My first guess would be a version of "loser pays" where it is presumed reasonable for a defendant to incur legal expenses in their defense up to 15% of the claimed damages, and if the plaintiff's claims are (mostly) rejected by the jury then they and/or their lawyers are on the hook for that. That would encourage the people with $20K in documented medical bills (etc) to ask for that and maybe a little bit more, but threaten a big downside risk if their larger claim is rejected. And note that in the first case, the defendant('s insurance company) will be paying out not just the $20,000 or whatever, but if they don't settle will also have to cover $3000 of the plaintiff's legal fees.
But there are lots of second-order effects and unintended consequences for anything like this, and I haven't put a lot of thought into it.
The victim can (and mostly does) get and use his own health insurance. It may be unfair to not have the party at fault pay for the costs, but it may be the lesser evil.
It's not clear to me from the linked Substack post whether we're talking about insurance for liability towards third parties (other drivers, pedestrians), or towards the passenger, or both. If it's about liability towards third parties, liability insurance is already required for all cars on public roads in most jurisdictions; there's no reason for the rules to be different between rideshares, taxis and private cars (but it seems like rules *are* different, with much higher requirements for rideshares).
If it's about liability towards the passenger, I see no reason to require liability insurance at all, or even have liability towards the passenger in the event of an accident. The passenger should be allowed to waive it even if the service provider would be liable by default.
After all, if you crash your own car, you either pay your medical costs out of pocket, or your health insurance pays them, or the liability insurance of the other driver does. When you use a rideshare, there's little reason to effectively require you to indirectly pay for extra insurance compared to that (via the driver's liability insurance getting factored into the price). In fact, if you already have health insurance, it sounds like you're paying twice to insure the same risk.
Theres also an opposing perverse incentive, where theres no point in fighting for more than the defendant+insurance can realistically pay, even if the damage is higher - and at 1E5, the additional amount you can extract from the defendant alone is often going to be small in comparison, and worth giving up for even modest increases in the odds of getting the payout. Note also that *this* effect is unavoidable for those harmed by it, while yours could be avoided by better control between the defendant and his insurance. The entire point of mandatory driver insurance was to avoid this problem.
I think the US does often have excessive damages, but the figure cited here is no evidence of it - how you interpret it entirely depends on what you already believe.
54. Note that those charts start around 1979, which was a local peak in criminality, creating a narrative of a consistent long-term decline in crime. It looks different if you start earlier and see the massive boom from from 1960 to 1980:
Interesting. Thanks for sharing the second graph. The late Kevin Drum hypothesized that removing ethyl lead from gasoline was the cause of the crime drop. I tended to prefer that hypothesis, but that graph would seem to undermine that theory. The moral of that story is always zoom out (on graphs).
Typo: "This problem may one day be solved by frontier labs may one day be able to change this" looks like two drafts of the same sentence accidentally spliced together, "may be solved by frontier labs" and "frontier labs may be able to".
9. I think I'm with DC here. If the argument is that settlements are too high, address the settlements. Otherwise it's a subsidy from victims of high-impact crashes, and the drivers/riders aren't paying for their externalities. It would be cheaper if we didn't require any insurance, but victims would be double-losers.
Legislation, probably? $100k seems really low. If someone's crippled by an irresponsible driver, that doesn't go far. Just arbitrarily reallocates costs.
To replicate this outcome consistently it would cap all damages at $100k. If that sounds crazy, then so should this. I don't know enough about law to give details about how it could work better.
TIL'ed weaboo/weeb is a word which should be odd given I live in a Far East Asian majority state. I feel this word has way less penetration that it seems; or I live in a bigger insular bubble than I thought lol.
The last part of #49 made me smile but honestly a bit much there, Musk may be a lot of things but a Nazi isn't one of them, not even neo. Sure he would have been had he been alive in Germany back then but so would most people.
Most murders aren't deliberate attempts to kill someone, it is usually using extreme violence to exert something and not caring if the person dies. Which morally and legally is seen as murder.
Gang killings are usually about exerting power in a dispute rather than deliberate assassination.
A murder is by definition a deliberate attempt to kill someone. Otherwise it is manslaughter. You seem to thinking of voluntary manslaughter where the perpetrator kills in the heat of the moment, using extreme violence without necessarily thining of the consequences, but not pre-meditated.
It depends on where you live. In California, second degree murder can refer to intentional but not premeditated killing, or killing someone with reckless disregard for human life.
They're saying most murders are the second one. The perpetrator isn't specifically trying to kill someone and ready to make sure they're dead. They're just inflicting general violence, not caring all that much whether or not someone dies.
#55: Car theft is another reliable crime statistic, for no matter how much they may dislike the police/think them useless, owners of stolen cars nevertheless have to report the thefts in order to cancel the registrations and, in many cases, file insurance claims.
Also, while better medical care may be saving some crime victims, the higher capacity of today’s semiautomatic pistols vs. six-shooter revolvers almost certainly means more multiple hits.
I'm skeptical that we've seen changes downstream of better guns in the last 30-40 years. The high-capacity semi-auto has been the generic "this is what you get if you want a pistol and don't have other preferences" since the mid-80s, and while there's some lag in what ends up in crime, it's not that long.
Yes, and that might have been relevant if I'd been talking about rifles instead of pistols. I was instead discussing the shift from a mix of revolvers and single-stack semi-auto pistols like the Colt 45 (7 round mag) to double-stack semi-auto pistols like everything on the market today.
That's a fair point then. I saw "high-capacity semi-auto" and thought that referred to a rifle and that you meant people were buying those instead of a generic pistol.
There's not going to be any statistics on the literal subject of "guns used in crime" because most crimes are e.g. muggings where nobody gets shot, nobody gets arrested, and the victim just says "he had a gun and it was bloody scary and I didn't get a chance to measure the damn thing". So you can imagine that all the muggers who aren't shooting anyone are using the crappy little Saturday Night Specials and I can't prove you wrong.
But we do have much better records on guns used in homicides over time, e.g.
Through the 1980s, the most common handgun used in homicides was a .38 revolver. And most of the rest were at least .32 caliber. From the 1990s, it's been mostly 9mm.
For historical context, the 9x19mm cartridge was invented in Germany in 1908 and gradually became the world standard over the next century, but it started in Europe and took a good long while to become popular in the United States. The so-called "wondernine" pistols that incorporated all of the features various European gunsmiths had been coming up with, came on the market in the 1970s and a few weirdos in the US thought they had potential but mostly people stuck with their .38 revolvers and .45 automatics. Mid-1980s, a number of major US police departments started using them and perhaps more importantly they started showing up in action movies in a big way, so after a few years of pipeline effects they became dominant.
That's correct about car theft, and applies to carjackings as well. In both instances the criminals are strongly incentivized towards taking newer/better/more expensive car models, and in turn the owners of same are more likely to have full-on car insurance coverage that is worth reporting a claim to.
The scenario of teenagers or gang members snatching a random beater just to have some yucks driving it around (a.k.a. "joyriding") does exist, but is statistically trivial.
Surely standard engine immobilisers are a huge factor as well? A 1980s car can just be hot-wired, a modern car will refuse to start without electronic contact with its own key.
Good point, although I believe there are a couple models of car where people were able to defeat that, spread the technique on TikTok, and then there were a spree of thefts of those cars. The manufacturers even had to distribute clubs to owners of that model.
Can you please post a link for the clubs thing? That sounds made up (if the owner sees the theft, there are probably better options). If it is made up, I don’t want to add to the search traffic for it.
> 33: Interesting as a way to build intuition for why Russia invaded Ukraine, h/t @MMJukic
(i got carried away and wrote a more general point)
I think that the war has made everyone slightly more interested in Russian politics, and I worry that people from certain backgrounds (e. g. Americans) may participate in the discussion with certain assumptions they don't even notice, that don't actually apply well to the Russian political system, and which they have a difficult time discarding.
E. g.:
1. I suspect it's very hard for someone from a more distributed political system to internalize the degree to which a country's political system can be *purposefully, cleverly, resourcefully optimized* to tune out any feedback signal, so that in its decisionmaking, it can fully follow its own preferences, instead of having to listen to annoying outside signals (like the country's own population, independent politicians etc.)
2. There's this myth of The People finally rising up against the tyrannical elite; which ignores that (if understand correctly) revolutions are mostly about whoever is currently in power already being very vulnerable to being overthrown or straight up agreeing to leave of their own will, e. g.:
— Nicholas II abdicated in favor of his brother, who then also abdicated; bolsheviks didn't overthrow the czar, they only overthrew the extremely weak interim government;
— Ceausescu signaled his weakness with that disastrous televised speech, lost support of the army after the death of the minister of defense, and was overthrown by his own fellows who already played a role in the Romanian political system;
— the Armenian revolution of 2018 involved mass protests, but it succeeded when Sargsyan left the office of his own accord, releasing Pashinyan from jail where he had been confined since the previous day.
In other words, if you have no ways of coordination because your institutions have been non-functional for a while, if you reasonably mistrust your peers (who knows if they'll report you to the secret police), and if your regime is united, coordinated, possesses many resources, has an army etc., the regime might get away with many things everyone hates and without facing any romantic glorious rebellion.
(1 and 2 are also are both present in the sentiment one sometimes hears, "most Russians must be supportive of Putin's policy, if they weren't, they'd overthrow him". I don't actually know if "most Russians" support Putin or not, it's really hard to measure when speaking out is dangerous; it's just not a locally valid logical step.)
3. the idea of Putin's role representing his country's interests in *some* way, e. g. acquisition of resources; while I feel it's more accurate to model his decisionmaking as trying to keep himself safe and keep his job. trying to keep some form of popularity is important to this, utter military failure signals weakness to elites (and signals are something to coordinate around! they create common knowledge in the form of "everyone knows that everyone knows that <...> Putin is weaker than usual right now"), while a "short victorious war" signals strength, etc.
I think there are more examples of this, but I'm not the best person to describe this; I'm not competent at political science.
I think that #33 may *somewhat* suffer from point 3 (but in general, they all sort of overlap.)
It gives the relevant vibe (that I think how Russian political elite — Putin and like the security council, whatever) that there are resources for which countries compete, and its peoples are an annoyance rather than the what countries are made of.
But also resources are not an end to itself: you need to control more in this world to be able to listen to less people inside and outside, so that no one can inconvenience you in any way when you make decisions they don't like.
Nicholas II was indeed overthrown by a popular revolution, assisted by mutinous army troops in St Petersburg. That is what is known as the "February Revolution" in Russia. The Bolshevik "October Revolution" was really a putsch and a takeover of the existing revolutionary government, as you say not really a popular revolution despite the Soviet propaganda.
Maybe what I said reads like "he decided to abdicate out of the blue", which is not correct, as there were gigantic protests, civil unrest, and basically mutiny.
But still I think my point sorta stands maybe? I mean:
· Nicholas abdicates,
· he's supposed to be succeeded by his son, a haemophilic child, but Nicholas abdicates on his behalf as well;
· the next supposed successor Michael Alexandrovich gives a weird neither-consent-nor-refusal (I will only accept if this-and-this referendum takes place);
· then everyone sort of decides that this was his way of declining the power;
· and then we're done, and Duma folks like Kerensky take it from there.
It wasn't *quite* the peasants-with-pitchforks situation, in the sense that it was ultimately other political actors who hastened the downfall of monarchy.
One contributing factor toward the success of the Iranian Revolution of 1979 may have been that the Shah knew he was going to die soon and didn’t really care about his country’s future.
Most Russians are broadly happy with Putin, and consider him a folk hero. 80% might be an exaggeration, but it's not all that much of one (this is judging by "self-initiative"*, which may be prone to mild exaggeration, but is definitely not going to mask "shy anti-Putin" folks).
*Say, who writes nice-nice things about Putin on their ballot, which is "mostly anonymous" -- this is initiative, and not something taken for social credit.
How many people would "write nice-nice things about Putin on their ballot"? It is probably not even 1%. Most Russians are a depoliticised blob which is neither for nor against anything and only wants to be left alone. The ones that consider Putin "a folk hero" are defo below 30%, I think, not "most Russians".
You may find other news of the sort just by googling "россия арест бюллетень дискредитация".
People casting a vote would be *realistic* to expect it's unsafe to write pacifist / anti-Putin things.
The point I am making is that "additional inscriptions on ballots" are subject to the same selection bias as Putin's approval rating polls; not only is the refusal rate high, but also the refusal *is more frequent* among those replying one way than the other. People were writing in anti-Putin inscriptions on their ballots in 2024 *despite* risking arrest.
2. Suppose we have somehow found stats about how this many ballots contain pro-Putin vs. anti-Putin inscriptions, and this many are blank. How does this generalize to the whole population? If you predict that most Russians are broadly happy, do you also predict that ballots with pro-Putin stuff are the majority — more frequent that ballots with just a check mark?
3. In any case, do you, in fact, have access to such statistics?
4. The "folk hero" interpretation is what the sequences would call a "burdensome detail". You think a majority of Russians not just approve of Putin, but consider him a folk hero *specifically*? What is this based on? Is that some statistical model I'm not familiar with?
***
I'm sorry if I'm being grouchy. I have spent a lifetime seeing Putin's system try to fake my consent to thinks I actively opposed, and I have had my chances to say fuck you to it, and in a way I am not surprised anymore. What is still beyond me is the phenomenon of buying this whole "he represents them all, that's what they are all like" story that Putin is specifically planting, and this being perceived as brave anti-Putinism.
I don't *actually* know what the distribution of opinions on Putin is like among Russians. But rationality is all we have in an uncertain world; if you care about what society's attitutes are like, it's also necessary to care if your inferences are valid and actually model the thing you think they model.
As someone said, "First they took our epistemology; we don't know what happened after that."
***
I worry I have taken this into culture-war stuff, and I apologize to Scott for endangering his comment section like that. Feel free to remove this comment if you feel it risks bringing in low-effort and/or hateful comments.
Well, to start out with, you have to start with discussing Putin as a Young WEF World Leader... who essentially backstabbed the "sclerotic Western European powers." I can give you more information about this, but... let it be said that there was a time in America when Sarah Palin was laughed at for saying "Russia is The Big Bad" and not very much later, Trump was being accused of "being run by Russia."
This was not an accident, nor was it happenstance, nor was Russia suddenly "a very convenient enemy." There were some very real enemies Putin made, and they were allowed to leave Russia (just, without the money they'd made).
Many Russians give Putin credit for stopping the West (see enemies made above) from thieving (this is at the heart of his reputation as a Folk Hero, you can think of this varietal of Folk Hero as like George Washington). This is above, and beside, Russian sentiment on any particular issue (importantly, right now, on the Ukraine).
<I'm agreeing with you -- my stance is merely that you can probably count "enhanced pro-Putin" ballots as a reasonable floor on "how many people Really Like Vlad">
I do not say that you agree with Putin (in all probability, your posting here increases the likelihood that you'd disagree with Putin...). And, indeed, what I've said above might be more accurate to characterize Putin as "czarlike" (aka personal popularity above that of Russia's government)
I listen to non-Russian sources whose job it is to know this... In so far as anyone can know the minds of over a million people, the data-gathering analysts can figure out who will win the next American Election (once people start deciding, anyway), and they can also figure out how well liked Putin is.
> I listen to non-Russian sources whose job it is to know this... In so far as anyone can know the minds of over a million people, the data-gathering analysts can figure out who will win the next American Election (once people start deciding, anyway), and they can also figure out how well liked Putin is.
This is a helpful clarification. Could you post a link to said sources? I'd like to evaluate them for myself as well.
> the data-gathering analysts can figure out who will win the next American Election (once people start deciding, anyway), and they can also figure out how well liked Putin is.
What I've been trying to convey is that, in my view, there is a deep sense in which predicting who will win the next election in the US is an easier task than figuring out how well liked Putin is inside Russia. At the very least it's a separate category of problems. I'd like to read more first, though.
A personal interview, on this subject at least (also where I sourced the kiwifarms anecdote above, and my source for most Russian/Ukrainian shibola).
I agree, in so far as you can find ways to evaluate "shy Trump" voters, but "shy anti-Putin voters" are likely staying shy for "very valid reasons" not just social opprobium. So it's hard to tell whether folks are "pallid Putin supporters" or "pallid anti-Putin voices" -- as the later have a very real incentive to behave like the former. Vehement anti-Putin voices are probably planning a revolution or something like that (and are probably tracked by intelligence foreign and domestic), so they're actually easy to find, as they have money-paths to foreign entities.
There's another point to be made: in a repressive regime, it is normal for people to "do what everyone else does" (so pallid supporters and opponents in some sense "are" the same -- and propaganda can actualize them into flip-flopping).
A counterargument I found for my own position while looking this up in this discussion is this discussion by Levada Center (one of the places where the 80% support figures come from): https://www.levada.ru/2025/02/26/k-voprosu-ob-oprosah/ arguing that refusal rates for their polling are not significant and are not significantly skewing the outcomes of their polls. I don't know enough to decide if I should buy this or not.
As a mostly-uninformed Westerner (The Berlin Wall fell before I was born but we hadn't scrubbed Cold War concerns from schools yet), the one salient thing to escape was a few years back now, where Russians were being arrested for holding up blank signs. I thought this was brilliant and easily exploitable for image macros, but I couldn't get a sense of how prevalent it was then, before, or since. Since has been the worst, with internet friends in Russia understandably not wanting to get deep into politics and broad disinterest from the media in "the two-week operation has now lasted one hundred weeks"
>What is still beyond me is the phenomenon of buying this whole "he represents them all, that's what they are all like" story that Putin is specifically planting, and this being perceived as brave anti-Putinism.
Let's also make regulations denying them visas and opening bank accounts, that will surely be helpful in defeating Putin
"Strongly disapprove" is also disjoint with "somewhat disapprove", which is not what I would call broad happiness.
"Somewhat disapprove" is not shown on either chart in the link provided, but the top chart does show "total approve", which has been topping out at 47% since last November, and sort of hanging around 45% on average, and is probably the closest thing to "broadly happy" for our purposes.
And again, that's Rasmussen, which has a reputation for showing more Trump-favorable results than other polls.
So, it's close-ish to half, but on the less-than side of it.
>that there are resources for which countries compete, and its peoples are an annoyance rather than the what countries are made of
Sure, and one of those resources is "glory", or "status" in rationalist-speak. I'm pretty sure that Putin's secondary motivation is to be a Great Russian Czar. He models himself on Stalin in pretty much the same way that Stalin looked up to Ivan the Terrible.
58: Regarding the startup idea, this falls into the classic trap of “assume that people are honest on dating apps”. Is there any reason why the optimal strategy would be to rate each option one’s honest opinion? I think the overwhelming majority of men would select either 0 or 9 on every swipe. I’m less sure about the optimal woman strategy, but my first guess is that they would use only 0 or 1 (why waste time on someone who doesn’t rate you a 9?)
This is true if a man wants to go on a date with a woman who isn't very interested in him. I can think of several women who I would be interested in dating if they were super into me, but not if they were only sort of grudgingly giving me a chance.
This is true in the abstract sense. In the real dating market that actually exists, there is no pool of single women super interested in the median single man. The options are go on a marginal date and hope to hit it off, or go on no dates. The binary “9 or 0” option dominates except at the very top of the distribution where men are just using the app for large numbers of quick hookups.
Yep, strongly agree, although I think the causation primarily goes the other way. Most women will rate most men very low (0-3) as they don't want to go out with guys who don't really care about them. Then men realise they don't get any dates unless they rate almost every woman highly, and decide to play the numbers game on any moderately good-looking woman. Then the ranking system ceases to convey any meaningful information and you end up with all the same problems as normal dating apps.
I'm gonna be honest - I don't really understand the reasoning process here. Can you elaborate?
I would intuitively think that really mismatched levels of interest would be bad - I don't want to date somebody who is really into me but I feel lukewarm about, because ultimately I'm probably going to disappoint them. And I definitely don't want to date anybody who would only accept because I'm so obsessed with them.
I definitely wouldn't want a big long-term mismatch, but I can see Scott's system making sense under certain conditions. It can be worthwhile to go on a date with someone you don't yet have strong feelings for, as long as you think there's a realistic chance that you might develop them. And if there are a bunch of people in that category, then rather than trying to date all of them, I might want to filter out the ones who are lukewarm about me and give it a chance with the few who are actively interested in me. (If nothing else, the odds of "I develop feelings and she retains them" are probably significantly higher than "we both develop feelings".)
This seems more like an argument for going on exploratory dates. Which makes sense to me! It's kind of silly we try to come to conclusions about how interested we are in people we've never met or barely interacted with. Although nobody seems like they're very interested in a process where you go on 3-5 dates with somebody and THEN decide if you're interested.
Yeah this partly depends on whether we're talking about an app for meeting strangers or an app for facilitating connections with people you already know. Scott said "I can think of several women who I would be interested in dating if they were super into me", so he might be thinking mainly of the latter case, but I'm not sure.
I think if we are talking about people who already know each other, it makes sense that there would be friction here -- like it or not, asking someone out does send a signal of interest, so you have to worry about embarrassment/harming friendships by creating awkwardness/wrongly giving someone the impression that you're into them when actually you're only looking for an exploratory date. And I can see the app helping with some of those things.
For strangers, I reckon you're probably right and it makes more sense just to lower your early-stage standards a bit. (Though I do think it's fine to settle on a 'no' after one meeting rather than feeling obliged to give the person a proper chance.)
You can’t ungamify dating. That’s like trying to ungamify poker.
I do think there is some possible algorithm that would incentivize an honest ordered ranking of potential partners. Enjoy your Nobel prize when you figure it out.
Econ Nobelist Ken Arrow discussed the problems of honestly ranked ballots in voting... but someone above suggested https://www.astralcodexten.com/p/links-for-february-2026/comment/210331449 just applying a curve to ratings so people who mostly give extreme ratings will have those shifted toward the median.
So the strategy here would be to give the median ranking to everyone except for someone you both like a lot and think you have a chance with. Sounds pretty good, expect that’s already the meta on like-throttled apps.
If the app is matching people with the top X% of their ratings, then a median rating could be lumped in like an extreme one. It does occur to me that people might be incentivized to avoid rating people at all if that could mess with their curve.
Surely the big problem is that a ranked ordering of people I've never met is pretty useless? Interest develops over time, after interacting with someone in person.
The Tinder algorithm actually seems optimal here, just a binary no-or-maybe.
Is there any analog to a Vickrey auction (winning bid pays second-highest bid price) that could be used in a dating context? ( Game theory is _not_ my strong point. )
I suspect there’s a way to do it where, instead of inputting a number, you do something like click higher or lower on the screen, and the app takes a bunch of your clicks and rescales them to give something like a uniform distribution across the numbers. Whatever gaming you might try, the app will undo.
I think this is pretty easy to fix from a design perspective: scale each user's responses so that their personal average is a 5, or give them a limited number of points to allocate across a set of 10 people each week, or etc.
#9, #48 It’s a talent to write as cogently as Cory Doctorow while being so cocksure.
Seems pretty obvious you’d need a lot of subsidy / VC capital yourself to undercut Lyft and Uber as a third wheel with no mindshare and no new gimmick (like autonomous cars)
It raises a few more problems, independent of what that company claims to have solved, such as wing erosion through cavitation and floating debris.
Also the side claim of revolutionizing military navies is also most likely just fanciful marketing speak - if anyone can disregard the economies of a technology, it's large militaries, and the large militaries that seriously tried hydrofoils found at most niche uses; improving the technology/economics won't change that. It's like making a sports car that's faster than a Ferrari and costs less than a bicycle, and the military still wouldn't have much use for it.
#55 "Aren’t many murders in very fast attacks where the murderer doesn’t get to choose how many shots/stabs to land?) and would welcome more research on this topic."
Probably important to distinguish between murder (planned) and manslaughter (more spontaneous, "In the moment"). No reason why a murderer shouldn't have enough time to contemplate such things, and in official/scientific crime statistics, I wouldn't expect those terms to be conflated.
To sort of defend theories that this might be militarily useful, we are talking about something a lot smaller and cheaper than previous submerged-foil hydrofoils, which could open up some new roles. They're talking about a 30 kt boat in Sea State 3, which is better than you can do with a conventional 30' boat. Put a diesel in it, and maybe there's some utility as an unmanned surveillance platform, particularly because it can now make normal warship speed. The problem is that I don't think Sea State 3 is enough. "WaterUber isn't running today because it's windy and the waves are too rough" is acceptable as an occasional thing, but it's probably not going to be enough for the military, particularly because it's going to happen a lot more wherever they happen to be. (It's 5 and over 40% of the time in the North Atlantic, which is the only number I happen to have to hand.)
We just need another Russian invasion of Iran to make the Caspian Sea militarily relevant; then these things (and the Caspian Sea Monsters) will finally have their chance to shine.
> 32: 60 Minutes recorded a segment on CECOT (El Salvador torture prison being used by Trump administration), then tried to suppress it (probably under indirect pressure from the administration), then changed its mind and showed it after all.
That is one way to frame the story. Here is a more neutral framing: CBS news hired a new leader (Bari Weiss) to oversee a news organization that lost trust by being rabidly anti-conservative. She noticed that a story about CECOT repeated claims that were at least debatable, at worst false. The reporters involved did not bother to try and understand the administration's position (for example, they made no effort to ask the administration why some choices were made). She told the reporters to at least ask the administration for comment, and refused to air their story until they did so. The reporters assumed that anyone asking for both sides of a story involving Trump must be a secretly working for Trump, and threw a fit.
> I have a bias towards Streisand Effect-ing things that get suppressed like this, so I’ll link it here even though it got on 60 Minutes eventually anyway.
Bari Weiss was a reporter at the NYT. She left (or was kicked out) because she saw nothing wrong with the NYT publishing an editorial by a republican senator (Tom Cotton) arguing that the BLM "mostly peaceful protests" were bad. She stood up for the idea that a newspaper ought to publish things, even if the reporters don't want their audience to hear those things. This was an unforgivable offense at the Times, and she was removed for it.
Her exit is solid evidence that she agrees with you that people should be allowed to speak, and the people who remain at the NYT believe the opposite. Scott, you are repeating a version of this story that cherry-picks facts so as to make a person who is advocating for fairness and reason into the villan.
I don't think "person believed Republicans should be allowed to speak in 2020" is solid evidence for "person is willing to criticize the Trump administration in 2025", and I think there are many, many counterexamples. I think she's amply displayed an anti-free-speech streak since then especially in her dealings with Israel critics.
Given that CBS has been trying to suck up to Trump ever since he gave them permission to do a controversial merger in exchange for (reading between the lines) being more pro-administration, I don't think it's a crazy theory to believe that the person hired to be pro-administration by their new pro-administration leader would quash an anti-administration story.
Can you present evidence for your claim that she told the reporters to ask for comment and they refused? What I see on Wikipedia is that "Alfonsi said Weiss had not contacted her before spiking the story and journalist Scott Pelley charged that Weiss had not attended any of five internal screenings of the story during the final stages of editing."
I said in the post that it was "indirectly suppressed", by which I meant Trump was tweeting that he didn't like CBS, everyone knew that the CBS merger depended on his good feelings, and so they complied in advance with what they knew his desires to be. I agree he didn't specifically send an email saying "cancel this".
They've also shown things that don't get the government's perspective, so I don't think "you can think of one 60 Minutes segment that doesn't fit this story" necessarily proves anything.
Only a full suppression or a complete redo would have benefitted the administration. Hard to see what it got out of the delay, other than an opportunity to comment which they apparently did not take.
Right. And the idea that Weiss only released it because she was "forced" (even though CBS announced that it would be aired later) implies there's someone both capable and willing to force her on that, like if David Ellison had a road to Damascus moment.
I don't think it's a completely crazy theory a priori, but CBS's overall behavior recently (including how they've covered other stories) looks a lot more like "trying to be more honest and rein in biased reporting" than "bowing to arbitrary admin pressure to bury stories".
Can you present evidence for your claim that she told the reporters to ask for comment and they refused?
"""
This article has the full text of the memo she sent about the story to people inside CBS (scroll to the end to see it). In that memo she points out several ways the report doesn't meet basic standards of fairness:
As evidence that she told reporters to ask for comment: Bullet point two from that memo:
"""
At present, we do not present the administration’s argument for why it sent 252 Venezuelans to CECOT. What we have is Karoline Leavitt’s soundbite claiming they are evildoers in America (rapists, murderers, etc.). But isn’t there much more to ask in light of the torture that we are revealing? Tom Homan and Stephen Miller don’t tend to be shy. I realize we’ve emailed the DHS spox, but we need to push much harder to get these principals on the record.
"""
As evidence they refused: The video was leaked. If the people involved took the feedback in that memo seriously, and wanted to address it, they would do so before showing the video to the world.
It is possible that the some video editor or IT person chose to leak the video without the reporter's knowledge, and against their wishes. Given the way this is being reported (search google news for "Bari Weiss CBS 60 minutes"), it is clear that the more common view among reporters is the one you shared: Weiss is biased in favor of Trump, and asking reporters to be even-handed is evidence of her bias.
Scott Said:
"""
Given that CBS has been trying to suck up to Trump ever since he gave them permission to do a controversial merger in exchange for (reading between the lines) being more pro-administration, I don't think it's a crazy theory to believe that the person hired to be pro-administration by their new pro-administration leader would quash an anti-administration story.
"""
Why do you believe she was "hired to be pro-administration by their new pro-administration leader"? It seems more likely to me that she was hired because the existing reporters let their politics color their reporting.
I can't read her mind. I can read her memo, and it seems to me to be a well-reasoned request to make the reporting less biased. Do you disagree? If so, what part is unfair?
> "On the afternoon of December 21, 2025, [three hours] prior to the airing of the episode, CBS News announced the CECOT segment had been pulled from the episode, and would air at a later date, without disclosing a reason. Multiple sources and commentators noted that this action was almost unprecedented, with stories almost never being pulled after such a vigorous screening process as typically occurs at 60 Minutes"
> "Alfonsi stated in an email memorandum to colleagues that the production team did request comments and interviews from White House officials, the State Department, and the Department of Homeland Security (DHS, which were all refused). And that the segment was reviewed five times and approved by the CBS legal department and Standards and Practices division."
> "Alfonsi had requested an interview with Secretary of Homeland Security Kristi Noem roughly a month before the feature was to be broadcast, but received no response; following additional requests for comment made to their office, DHS referred all questions to the Salvadoran government. Axios has reported that the State Department provided comment as well. In a response sent on the 18th, the Trump administration questioned why 60 Minutes had not focused their efforts on the experiences of Angel Parents, "whose innocent American children [had] tragically been murdered by vicious illegal aliens that President Trump are [sic.] removing from the country."
> "'Government silence is a statement, not a veto. Their refusal to be interviewed is a tactical maneuver designed to kill the story,' she said."
> "At the meeting, 60 Minutes correspondent Scott Pelley again brought up the fact that Weiss had not turned up to any of the five internal screenings of the segment during the final stages of editing. Pundits have noted that her intervention came only after the piece had been finalized and promoted on social media."
So the impression I get is that the people writing the story tried to get both sides and comments from government officials, they presented the story five times in a way that Weiss normally would have had an opportunity to see and comment on, they got it reviewed by the procedure for reviewing these things to make sure it was fair, at each step everyone agreed that since the government officials had refused to comment, it was fair to air it without their comments, since they didn't want to make the government able to trivially crush any story by refusing to comment on it. Then three hours before it was broadcast, Weiss, who should have been monitoring this entire process, suddenly cancelled it in a way that had basically never happened before, saying it needed to get a government official's comment. So either Weiss is extremely incompetent or she was acting on some sort of political orders.
But I've edited your comment and the resulting thread into the main post so people can read it and get a more thorough view of the situation.
I’m not sure how reasonable it would be for the head of CBS News to be present in preliminary internal screenings of 60 Minutes segments. Definitely weird to pull such a hard-hitting segment on short notice, but if Weiss thinks that the internal review departments are incompetent, then her overruling their decision to green-light the segment isn’t necessarily evidence that she was pressured, just evidence that she has different standards than them.
I am preregistering my bias as generally anti CECOT and Trump skeptic, fwiw.
Nevertheless --
I believe the most plausible story accounting for somewhat charitable readings of all sides is that she did not like the story, but that her memo was still some rough approximation of the truth, and she wasn't conspiring to be responsive to trump's whims while lying about it.
I have high uncertainty whether she didn't like the story because, as she said, she did not think it broke new ground relative to other existing reporting, or because she has an arbitrarily high bar for all stories against the administration, or because of some middle ground where she has an arbitrarily high bar for stories that pile on issues already litigated by other outlets. (I find the last of these most likely because in my experience people instinctively soften bad news a *moderate* amount, but not typically to the point of outright lies, so both extremes of perfect honesty and maximal deception seem less probable to me, and this was the only middle ground I was clever enough to imagine.)
She clearly has communication problems with her staff, though no wonder, fostered by the perception that she's an outsider brought in to "correct" their political views, which are no doubt sincerely held, or even personally interpreted as objectivity amidst a hostile political environment.
I downrate the idea of a premeditated spiking because that is inconsistent with her lack of earlier involvement when it would have been easier to inject delays, inconsistent with her lack of apparent control over leaks of the story or leaks about the story's spiking, and inconsistent with her claims in the memo that she would help them get the story over the bar by providing personal contact info of other officials.
Offering to help is a form of expensive signalling. If someone takes you up on it and you do not offer any meaningful assistance it raises the likelihood they will move you from "trembling hand" to "spiteful defector" in their assessment of your behavior, particularly if iterated. I understand that it's possible her game theory is stalling for time, that is plausible but seems less likely to me than her behaving like a normal human in a contested environment.
I do not have good baseline information about how likely it is for a person in her position to watch a story five times in preliminaries. Maybe she was seething with rage at the very idea of watching the story she hoped someone else would spike so she wouldn't have to. Maybe her position doesn't usually do all those reviews. Maybe she's new and should have been there but doesn't actually realize that's a line level expectation (possibly exacerbated by the friction of her presence which are preventing open upward feedback).
One of the most interesting pieces of evidence in the aftermath is that nobody appears to have been fired over this. That raises my p() that they are an organizational unit with many strong opinions but genuinely trying to muddle through from both sides, and lowers my confidence in the idea that they have crossed some kind of us or them rubicon internally. (Though... I'm not sure this is fair. I still expect someone to be let go / pushed out in the next six months after enough time has passed it seems unrelated to this and not sure from which side. Would that make me more convinced this was actually a gutter fight where everybody was playing dirty? Maybe it should proportionally, hmm.)
The facts you are presenting are an account of what happened by people who are not disinterested parties. We have no way to figure out if these claims are true, or are the whole truth. It comes down to the credibility of the people involved. That will never give us proof, but it is the best we can do with limited information.
Scott, as I read your response I think of your essay on "Trapped Priors": You seem to start with the assumption that Weiss is biased in favor of Trump, was installed because she is biased in favor of Trump, etc. I don't share that assumption. I think she is an old-fashioned liberal (to the left of 75% of the American public). I think the reporters are the new sort of democrat, the kind that does not tolerate dissent from the party line (to the left of 95% of the American public). I can't know for sure which of us has the wrong prior.
To me, the following factors weigh in favor of Weiss:
1. On at least one occasion, Weiss stood up for the principle that a news organization should allow editorials that the reporters do not agree with. She did this even when it cost her a job. Very few people can say this. That tells me she has some integrity. It is possible that she was a secret republican mole who went to work for the NYT so she could get canceled to fake credibility, but I doubt it.
2. The memo Weiss wrote is well reasoned and seems to be a genuine attempt at making the reporting fair and honest.
3. The reaction to that memo by journalists generally makes no effort to consider anything she actually said. If the reaction was along the lines of "I can see her point about ..., but I disagree because ..." then I would be inclined to think the reaction is a serious attempt to get a good result. I don't see that. At all.
4. I watched the leaked piece. The issues Weiss raises are, in my opinion, real issues that should have been addressed.
5. I have worked in orgs that were failing, and had a new manager put at the helm. The people from the failing organization hated the new manager, and did everything possible to undermine her. I bet the long-term CBS reporters HATE the fact that someone kicked out of the NYT for not being far-left enough made her own news organization, beat them at the job of producing reputable news, and now has the gaul to tell them what to do.
Isn't it undisputed that the journalists tried to get the administration to offer their side of the story, and the administration refused to do so? That would seem to entirely negate the force of the objection that it would be irresponsible to run the story without getting the administration's side.
My impression from this is that the previous attempts had been checkboxing "we (hostile media org" CCd an admin official knowing they don't like to talk to us", and Weiss said "no, I want you toake a real serious effort to get their side of the story here".
1. I still don't think this proves anything. The Communists protested when people suppressed Communism, but that doesn't mean they were generally free-speech loving people who practiced free speech when they were in power.
2. I agree the memo is the best evidence for Weiss' side, but good writers can write memos approximately that well-reasoned and genuine-sounding for basically any scandal you give them. It's not hard for a writer of Weiss' talent to write a memo that says "I was right and doing reasonable things".
3. I think the journalists' response to the memo is "She says she cancelled because we didn't get an administration interview, but we tried to get an administration interview and they refused, and there are known procedures for dealing with this and she ignored all of them in an unprecedented way". I think this is a strong response.
4. Fair enough, I didn't watch the piece.
5. I think this requires that we throw out the testimony of everyone except Bari because they might be biased, while also trusting that Bari is the single unbiased individual in this entire affair (even though of course she comes into this fraught situation with biases of her own). I agree there's a lot of uncertainty because everyone has reason to spin things in their own direction, but I don't see a reason to trust Bari more than anyone else.
But fine, you've convinced me there's some possibility it wasn't politically motivated, though I still think <50%.
Is Bari Weiss analogous to those communists? My understanding is that she has called herself “center left on most things” and “left-leaning centrist”. The Free Press under her leadership has been willing to criticize the Trump admin: https://www.thefp.com/p/is-donald-trump-breaking-the-law
Weiss’ previous integrity is definitely impressive. But between them and now, she worked out an agreement where CBS bought her company. I don’t know much about the integrity of most people in that position, but the most salient example I can think of is from Australian 2010s politics. The opposition leader Malcolm Turnbull resigned when he realised he couldn’t lead the (right-wing) Liberal party to act on climate change. But later, he came back into office by agreeing to give up and accept the party’s anti-climate-action position.
Possibly this is just the media I see. Do you know of stories where people made these principled stands, left, then came back to an organisation which still rejected those principles, and managed to implement the principles anyway? I’d like to see those stories, for their own sake.
> On at least one occasion, Weiss stood up for the principle that a news organization should allow editorials that the reporters do not agree with. She did this even when it cost her a job.
Bari Weiss resigned from the NYT. She wasn't fired or kicked out for standing up for the Tom Cotton op-ed.
I'm banning you for saying false things in any overly confident and hostile tone. For my response to what you're saying, see my discussion with Frange above.
I think a lot of the issue people are taking here is that the contentious framing (which I maintain is inaccurate, as Frange stated) comes across as selective bravery. A lot of influencers in this sphere are happy to 'stand up to' people so long as the people they are standing up to aren't the people who tend to retaliate harshly when people do so, and that creates all the wrong incentives.
>I have a bias towards Streisand Effect-ing things that get suppressed like this, so I’ll link it here even though it got on 60 Minutes eventually anyway.
This doesn’t sound like a neutral framing. CBS was only losing trust in the general sense that all legacy media organizations have been losing trust as people move to distributed media. There was no sentiment among anyone that CBS was “rabidly anti-conservative”.
Bari Weiss had a very prominent career in between the New York Times and CBS, and understanding what she was doing in that career is relevant to understanding why she might have gotten hired at CBS.
I would think that it’s only rightists who think that, not most non-leftists. But even they don’t think CBS was worse than ABC or NBC (let alone print media).
Trump singled out CBS in particular as being biased against him during the campaign, after 60 minutes ran that interview with Kamala Harris that was edited to make her sound better. Trump went off on how biased and bad CBS is many times, publicly, in campaign speeches to his voters. He sued, and CBS ended up settling. So it wouldn't be unreasonable for someone, even if they're a centrist swing voter, to get the impression that CBS has an anti-Trump bias of some kind. It was a narrative that was going around.
Ah interesting! I hadn’t noticed that he was discriminately attacking a single media organization rather than just indiscriminately attacking lots of them!
I gotta say that the "neutral framing" here is very favorable to Weiss, and not entirely accurate.
Cotton's op-ed was not that the protests were bad, it was that the military should be deployed domestically to clamp down on them. It was literally (and accurately) entitled "Send In The Troops".
(Also I honestly am not sure why you feel like this is the main reason she left the NYT, rather than one of many she herself cited. https://www.bariweiss.com/resignation-letter)
But I'll give another reason why the NYT shouldn't have published it: because it's not 1995. Tom Cotton can publish anything he wants at no cost and be confident it will be read by a decent number of people. The NYT doesn't need to signal-boost his opinions if they don't think they have merit, because the days when something needs to be in the NYT to be read are long dead. They should have an ounce of selectivity about which things they print.
Also putting scare quotes around "mostly peaceful protests" is kind of pointless. The protests after George Floyd died were MASSIVE, with tens of millions of participants. They can both be mostly peaceful and, in aggregate, cause a lot of damage.
Also, framing Bari Weiss as a free speech warrior of some kind is just not credible. She's been hounding people for wrongthink on Israel since she was in undergrad. Weiss (like mos people, honestly) talks about the importance of free speech for ideas she supports or agrees with, and turns around and suggests that ideas she doesn't like are unworthy of consideration.
For those interested, the New Yorker recently did a profile of Bari Weiss, including a discussion of the 60 minutes controversy. I like the New Yorker, and I believe their stories to be well-fact-checked, but I understand if others feel they are biased. Regarding this story, they don’t add much beyond what Scott said here, although it does mention that she had offered some earlier suggestions to the story that were incorporated, before later cancelling/deferring it, which weakly supports the idea that the cancellation could have had outside/political motivation: https://archive.ph/4gWia
“””
A few days before Christmas, “60 Minutes” was set to air a report on cecot, the prison in El Salvador where the Trump Administration had sent more than two hundred deportees in March. The men, most of whom were from Venezuela, had been spirited out of the U.S. on a series of late-night flights, in violation of a federal judge’s court order. Sharyn Alfonsi, a “60 Minutes” correspondent, had interviewed two of them. The stories were harrowing. “There was blood everywhere, screams, people crying, people who couldn’t take it and were urinating and vomiting on themselves,” one of the men told her. She had reached out to the Trump Administration for comment but received only a cursory, two-sentence response.
Like all “60 Minutes” stories, the cecot segment had been fact-checked and vetted by the network’s legal department. Five separate screenings were held for various editorial stakeholders. Weiss was supposed to attend the final screening, on Thursday afternoon, but she had missed it. She didn’t see the segment until late that night, e-mailing suggestions for a few changes that were incorporated into the piece. The network promoted the segment as the lead story for that Sunday’s episode. Alfonsi flew home to Texas.
The cecot story was being finalized almost exactly as Ellison was pursuing another major expansion of his growing media empire. Earlier that month, he had launched a hostile takeover bid to purchase Warner Bros. Discovery, the film-and-television conglomerate that had already announced a deal to be acquired by Netflix. In many ways, the outcome depended on Trump, since regulatory approval would be required for either sale to go through. “I’ll be involved in that decision,” Trump had told reporters.
For Ellison, this was suddenly a problem. “60 Minutes” had recently aired an interview with Marjorie Taylor Greene, a former Trump ally who had fallen out with the President over his resistance to releasing the F.B.I.’s files on the convicted sex offender Jeffrey Epstein. “they are no better than the old ownership,” Trump posted on Truth Social after the interview aired. “Since they bought it, 60 minutes has actually gotten worse.”
On Saturday morning, Weiss, who reports directly to Ellison, told “60 Minutes” producers she was concerned that no officials from the Trump Administration had been interviewed on camera for the piece. Specifically, she wanted to include the Administration’s argument for its use of the Alien Enemies Act, an eighteenth-century law that Trump officials claimed allowed them to deport Venezuelan migrants to El Salvador without due process. “We should explain this, with a voice arguing that Trump is exceeding his authority under the relevant statute, and another arguing that he’s operating within the bounds of his authority,” Weiss wrote in a note to the producers. “There’s a genuine debate here.” She had “tracked down” numbers for Tom Homan, the border czar, and Stephen Miller, Trump’s chief immigration adviser, which she sent to the “60 Minutes” team.
On Sunday, three hours before broadcast, the cecot piece was officially pulled from the lineup. “My job is to make sure that all stories we publish are the best they can be,” Weiss said in a statement. “Holding stories that aren’t ready for whatever reason—that they lack sufficient context, say, or that they are missing critical voices—happens every day in every newsroom.” Alfonsi felt betrayed. “In my view, pulling it now, after every rigorous internal check has been met, is not an editorial decision, it is a political one,” she wrote in a note to her colleagues. “Government silence is a statement, not a veto. Their refusal to be interviewed is a tactical maneuver designed to kill the story.”
The next day, the Ellisons announced a significant sweetening of their bid for Warner Bros.: Larry would personally guarantee $40.4 billion of the funding. Weiss didn’t come to the CBS offices that day. She joined the morning’s daily editorial meeting via Zoom, beginning with what seemed like a rebuke of Alfonsi. “The only newsroom that I’m interested in running is one where we are able to have contentious disagreements about the thorniest editorial matters and do so with respect, and, crucially, where we assume the best intent of our colleagues,” Weiss said. “And anything else is absolutely unacceptable to me and should be unacceptable to you.”
That afternoon, during a “60 Minutes” staff meeting, Scott Pelley, a longtime correspondent, expressed frustration that Weiss hadn’t attended any of the screenings of the segment or communicated directly with Alfonsi. “She needs to take her job a little bit more seriously,” he said. A former CBS staffer was soon circulating an open letter to Ellison, expressing alarm at “a breakdown in editorial oversight” that risked “setting a dangerous precedent in a country that has traditionally valued press freedom.” A former CBS executive told me that, even if Weiss’s concerns had been valid, her decision to cut the segment at such a late hour had opened her up to charges of corporate interference: “It makes you wonder, Did someone call once they saw the promo on the air and then she spent more time on it because there was some big complaint?” Sources close to Weiss and Ellison said that Skydance leadership had zero involvement in the story and did not screen the piece.
These sources also told me that Weiss “readily realizes and admits that she was not as knowledgeable as she should have been about the timing of the marketing and promo process at ‘60 Minutes.’ She brings the sometimes chaotic energy and work ethic of a startup, but she also realizes she needs to work on having more executive discipline.” Weiss also seemed to be struggling with the fact that, at a time when the Trump Administration is routinely lying to the public and straining to justify blatant abuses of executive power, often with violent or deadly consequences, she was still wedded to the idea of news coverage as a contest of ideas, in which both sides of the debate are equally valid. Privately, she has expressed alarm at many of the Administration’s actions, a person close to Weiss told me. But, in her role as the editor-in-chief of CBS News, her main concern is being able to book its main players on her network’s shows.
The dynamics of the Manifold COVID origins market are explainable based on three factors:
(A) Market structure that'll likely never resolve, leading to only a small number of motivated individuals dominating the trading; (B) financial changes in Manifold affecting the behaviour of those individuals; (C) user churn among them
(A): As you note, the market structure is such that it is very unlikely to resolve (ever).
This means that people who I'd consider to be "serious" traders don't really participate much in this market on a volume basis - it's just not clearly profitable when your mana will be locked forever and price isn't really tethered to reality by an outcome happening.
The market has had 4.6m in mana volume; about 25% of the total is Peter Miller alone on the zoonosis side and about 15% is the top 3 traders on the lab leak side.
In contrast, the biggest trader by position I'd consider 'serious' I can see in the top lists has about a 16k zoonosis position, or 0.34% of volume.
When I say the top holders aren't "serious", I don't mean this as a diss to them personally, just that they're not necessarily the most profitable users on site, and they don't trade a lot on other unrelated markets.
(B): Manifold keeps on iterating on its financial system, in particular the loan system. Over the last year or two they've removed and recreated the loan system a couple of times, as well as adding per-market leverage caps and reducing loan dispersion rates. This means that there's less liquidity flowing for single-issue traders to stake on this market; and when there are relatively few major participants, this matters a lot.
(C): Some of the biggest holders on the lab leak side have simply lost interest. "Mike P", the biggest holder on the lab leak side, used to trade ~daily and appears to have quit Manifold in early 2025; "Robert Sutherland", the second biggest holder, traded frequently through April and now only does sporadically, I'm guessing substantially due to the loan system changes.
Basically, the whole market is "Mike P, Robert Sutherland, and George Riddick (lab leak) vs Peter Miller (zoonosis): who will blink first?" and so far it seems that Peter Miller has more staying power.
I think the dynamics of this market are relatively unique -- most markets on Manifold don't have this particular issue, and it doesn't apply at all on other forecasting or prediction market platforms, where there's a strong editorial voice ensuring markets are in principle resolvable.
The part that is most transferable is looking at forecaster P&L for top traders on a question in a real-money market, although I'd note that that's also not without thorns and shouldn't be taken naively.
E.g. it's very common for sophisticated traders to hedge across correlated markets, so just because you see a profitable user taking a big position on one side of a particular question, does not mean that they necessarily strongly believe that side.
I'm not sure if I understand the question you're asking.
If it's literal, "how can this market resolve?", I guess it could in theory if a lab leak smoking gun was discovered, or if a complete zoonotic transmission chain were documented in a high status scientific publication.
If it's emphatic, "why do people trade at all when obviously it won't resolve?", idk, trading is fun and the debate was spirited
If it's technical, "how could one write a prediction market like this that would resolve?", you can ask about a proxy and specify a cutoff time, e.g. will a paper be published in a top tier journal that concludes one way by date XYZ.
It could be vague, but I was curious about what was actually written. So I looked it up and here is the section I found relevant:
This market resolves once we have a definitive answer to this question. (i.e. "I've looked at all notable evidence presented by both sides and have upwards of 98% confidence that a certain conclusion is correct, and it doesn't seem likely that any further relevant evidence will be forthcoming any time soon.")
This will likely not occur until many years after Covid is no longer a subject of active political contention, motivations for various actors to distort or hide inconvenient evidence have died down, and a scientific consensus has emerged on the subject. For exactly when it will resolve, see When will the Covid lab leak market resolve?
I will be conferring with the community extensively before resolving this market, to ensure I haven't missed anything and aren't being overconfident in one direction or another. As some additional assurance, see Will my resolution of the COVID-19 lab leak market be controversial?31%
(For comparison, the level of evidence in favor of anthropogenic climate change would be sufficient, despite the existence of a few doubts here and there.)
If we never reach a point where I can safely be that confident either way, it'll remain open indefinitely. (And Manifold lends you your mana back after a few months, so this doesn't negatively impact you.)
Hi, I have the third largest zoonosis position on that market. I agree with you directionally, but also want to push back a bit on your framing.
For example: My position is indeed much smaller than Miller's and I don't think of myself as a serious trader (proof: serious trader should trade on other markets). But I'm not a noob either — e.g. see my profits/calibration https://manifold.markets/VaclavRozhon/calibration . I just care more about testing my calibration than winning mana. I can see more semiserious people with semiserious stakes in this market. So there should be some signal there.
Then again, I of course mostly agree with you and don't really want to claim that this market is somehow efficient because I don't think it is -- if serious people traded, I think lab leak would have crashed much sooner.
I appreciate the pushback. I admit I didn't really rigorously check the whole list for 'seriousness', I was mostly just looking at the usernames to see who I recognised, assuming I'd recognise more or less every 'serious' trader. But you actually have more lifetime profit than I do! So, clearly, this method is imperfect.
FWIW, you're at least half right. There was a bidding war for a while. The market crashed after Scott's post about covid origins, some large users like Joshua started asking why not just take it down to 10% to match Scott's views. Mike P started buying mana to prop it up, I think he spent ~$300 on that until the No bettors gave up.
I had previously steered clear of any large bets in the market, but at some point I decided to just see if I could overwhelm Mike and I got it back down below 50%.
I mostly gave up after that contest -- you can see my average price in the market is 51%. I've thrown a few smaller bets in below 50%, but not so much compared to fighting around the middle. If it was just me, I probably would have just thrown in all my mana as a wall at 49% and called that good enough.
Other people have since decided to take the market well below 50%. I don't know their motives, or what they are expecting as far as resolution. That could either be a shift in sentiment against the lab leak theory or a shift in the average participant in the market.
FWIW, I think the actual odds that Covid is a lab leak are non-zero but less than 1%. Others disagree, and I think the market has little connection to any real assessment of the odds, either way.
I'm not even sure what the rational price should be. Should it match expert surveys and settle somewhere in the 10-20% range?
Even if it were certain to resolve as No in 2040, what should the price be based on the risk free rate? People will bet on the UFO market at 95% for a guaranteed win in ~2 years, so should it be 43% return for 14 years? Or maybe less because it could potentially resolve sooner?
Worth emphasizing: The market dropped after *Scott's post* about the Rootclaim debate, not after the Rootclaim debate itself. This, I think, proved that the market probability was not being determined by informed actors.
What’s the mimetics of weaboo? Terms don’t escape containment unless they have a reason to spread, and weaboo doesn’t seem particularly clever to the average folk without context. Is it because it’s fun to say and vaguely Japanese sounding?
It's because wapanese/weeaboos/weebs/whatever are a real phenomenon, significant enough to be regularly discussed in certain circles (basically anyplace anime/manga-adjacent, for starters). which means there's strong incentive to find a short memorable name. The first candidate was obvious and descriptive but sounded like it might be sort of racist, so as soon as there was any significant usage of the new "means the same thing as wapanese but Officially Not Racist" term, it took off. I had not known that 4chan was the source of this linguistic evolution, but it makes sense.
And the meme continues to spread. We now also have "Wehraboo" (from "Wehrmacht"), for people who are as gushingly enthusiastic about WWII-era German military culture as the Weebs are about Japanese culture.
We do not have "Weeaboo II: Electric Boogaloo". Yet.
The first time I heard the term Koreaboo, I asked the person who used it if it meant a white person obsessed with Korea, or a Korean person obsessed with Japan. He told me it meant the former, and added that the word for the latter was "traitor".
Before you know it, we'll have so many people gushingly enthusiastic about calling out other embarrassingly enthusiastic fans that we coin "Electric Boogaliboo".
Regarding item #10, I'd like to recommend the book *Wild Geese Returning: Chinese Reversible Poems*, by Michèle Métail (a member of the Oulipo, for those who know what that is), originally published in French in 2011 and translated by Jody Gladding in 2017. It's about the whole genre of poems that Su Hui kicked off, which persisted for centuries. About 1/4 of the book is about Su Hui's poem, including some translated sections (as far as I can tell, the whole poem hasn't been translated; but as far as I can tell the project would be practically and maybe theoretically impossible). And then it includes a whole lot of other reversible poems. A great book.
>The only crime that isn’t like this is murder - everyone notices a missing/dead human, and the police have to investigate all of them.
One could imagine a dynamic where the police itself becomes lazy and uninterested, and would rule potential homicides as accidents or suicides to save themselves from the hassle of investigating it.
Or because their funding/promotions/etc rely on hitting certain metrics.
I basically don't trust any statistic from a police department for the same reason I haven't trusted standardized test scores since No Child Left Behind: the measures became metrics, there's overwhelming incentives to fudge them and little oversight to catch it.
Adjacent example on the BBC today. Police assumed that a dead body they found was non suspicious (they thought he'd fallen). Only later, days later, did they realise he'd been shot with a shotgun.
There was a similar example I saw on some TV show a while back, where it appeared a man died alone in a hotel room without any injuries, but during the autopsy, they found out his heart exploded. Turns out he got accidentally shot in the crotch from the room next to his, and the bullet entered in such a way that it closed the wound behind itself.
38. Yikes. Here’s an excerpt from the description:
“If your ancestors aren’t entitled to control your beliefs and practices, why are you entitled to control the beliefs and practices of the strangers who surround you? The best cultural policy is not cultural preservation, but cultural competition.”
I’ve seen Bryan Caplan’s view on this stuff, and I’m not sure I 100% understand it, because at face value it seems obviously wrong. If a culture of cooperation meets a culture of defection (I.E. Welfare being generally used as-needed vs. welfare being maximized and gamed to produce the most money for the least amount of work) the defecting culture will “outcompete” the cooperating culture.
In liberal democracies, with welfare, it seems impossible for the open-borders world Caplan advocates for to function. Population equates to political power, and there is little incentive to *not* game welfare when you have no preexisting social bonds to the society you’re receiving welfare from.
I feel like this was a much broader instance of AI-trickery, since that Uber Eats thing was only interesting to a specific corner of the internet, while probably half of America saw the Maduro pictures on the news and thought they were real. Of course the impact was smaller, since it wasn’t really misrepresenting the situation at all.
1. Is it actually true that modern welfare states are "cultures of cooperation"? Or do they just have strong institutions that make it hard to cheat? ("Welfare queens" have been a stock boogeyman for Republicans for decades, regardless of what the immigration rate is.)
2. Is it actually true that the "social bonds" you get from being native to a society disincentivize antisocial actions? If so, why is the crime rate for immigrants *lower* than for natives?
(Also, while "social ties discourage defection" is an argument that makes sense for not doing crimes to your friends and neighbors, I'm not sure the same argument applies when you're defrauding the distant and impersonal federal government.)
3. If an immigrant plans to live in their new country forever, don't they have the same incentive as any native to ensure that the country they live in stays strong?
2. Is it actually true that the "social bonds" you get from being native to a society disincentivize antisocial actions? If so, why is the crime rate for immigrants *lower* than for natives?
Filtering and a higher baseline crime level would be a strong culprit if you're talking about the US. In Europe migrants are significantly overrepresented in number of crimes.
3. If an immigrant plans to live in their new country forever, don't they have the same incentive as any native to ensure that the country they live in stays strong?
Not if gaming the system benefits you personally. There is a very strong incentive to take free money for you personally, even if in some abstract and minuscule sense contributes to harming the nation.
---
But take this to the extreme as an example. To be eligible for SNAP in the US, a 3-person household needs to make less than $20,352, meaning that ~95% of the global population would qualify for basic US welfare, not even considering benefits that extend well above the $20k maximum (larger households, housing assistance, etc.).
Living on welfare in the US doesn't give you a great quality of life compared to what you can get by working. If you're able to, most people work. Living on welfare in the US compared to much of the world gives you a great quality of life, and you also don't have to work. Open borders, with the sort of welfare systems we currently have seems incompatible to me.
And the cultural argument seems obviously wrong too. If a hundred fundamentalist Muslims move to your town and have 5 children on average, while your population has 1.5, over a few generations your community will become Muslim. Sure, there will probably be some defections over to Western Liberal society, but in the meantime you will get significant conflicts of values that can cause serious problems, and in the long-term your cultural values may be less "competitive", so they lose.
And I'm a very pro-immigration person. I think that for persons with the means to sustain themselves the bar should be basically "Show up with a clean criminal record, and you're given a SSN." But then Caplan comes in with his advocacy for complete open borders and I feel that my (IMO internally consistent) view becomes anti-immigrant. IK Caplan is a libertarian and doesn't support the welfare state in general, but his advocacy for this single issue, without clarifying that it probably wouldn't work unless we overhauled a large part of the rest of society as well, is pretty bad.
It should probably be pointed out that the welfare state isn't the actual issue. The reason for migration in the current age (and, to be fair, in the past as well) is overwhelmingly the Bank Robber's Maxim: that's where the money is.
If we stop handing out money, in the form of welfare, that does not necessarily stop the immigrants from coming - not if they see a different way of effecting a move of money from someone else's pocket into theirs (hint: I am not talking about hard work).
Though coming to America and working hard is in fact an excellent way to move money from someone else's pocket into your own. Often much more effective than staying in your home country and working hard.
Absolutely! However, this assumes that there will be a job for you, somewhere. Currently, that's still a decent bet in the US. In Europe, the situation isn't as rosy.
Not as rosy, but here in Germany I see many job-openings, many remaining vacant - and the hard work done mostly by immigrants. Immigration into our labor-market Dubai-style would be just as great here as in Dubai. As immigration into our social-security is often easier as into our workforce ... well, 'show me the incentives; I show you the outcomes to expect'
I've always sort of assumed, without knowing for sure, that libertarians in favor of open borders aren't actually one-issue but rather have decided that open borders is the only chance we have of getting rid of the Welfare State, since the two cannot coexist, and consider on balance that it's worth the turmoil that would ensue. I'm kind of with them up to that very last step but can't quite make myself go all the way.
If we go that route, we need to get rid of the welfare state first. The problem is that politicians won't want to do that because maintaining the welfare state gets you votes, especially if you bring in new welfare recipients to vote for you.
Sure; that's why people like (I assume) Caplan don't believe you can get rid of the welfare state first. The only way is to make it so untenable that the incentives for the politicians push them in that direction. Like I say, I can't disagree, as we've never had any real success cutting it back even marginally -- how the founders would have laughed at the suggestion that we would spend more on old-age pensions than on the military! I'm just afraid that the downsides of open borders are too terrible to accept even if it allows us to get rid of the welfare state.
That is certainly not Caplan's position. While he is very critical of welfare-stateism, he is ardently pro-open-borders. He even refutes Friedman's thesis (welfare and open-borders: incompatible) in his (great) book "Open borders", Caplan there : We let (our) babies in. Into our welfare-state. Without being sure they will not turn into welfare-trash. Ergo: No fundamental incompatibility.
"If an immigrant plans to live in their new country forever, don't they have the same incentive as any native to ensure that the country they live in stays strong?"
Yes, but an unassimilated immigrant also has the incentive to transform the country they now live in., to something culturally similar to where they came from. These two can conflict, if e.g. the culture of their ancestral homeland wasn't conducive to national strength. And pretty much by definition, their ancestral homeland wasn't strong enough to hold on to them.
Traditionally, most immigrants haven't had the power to do anything really transformative along those lines, and had to settle for maintaining enclaves of their home culture for a generation or two. But as we move into an era with large-scale migration coupled with the attitude that assimilation is problematic because something something settler colonialism, salad bowl rather than melting pot., we should want to watch carefully to make sure that's still the case.
> Traditionally, most immigrants haven't had the power to do anything really transformative along those lines, and had to settle for maintaining enclaves of their home culture for a generation or two
Lots of examples where they have though, like European immigration to the Americas, Australia, New Zealand. Or the Chinese in Malaysia and Singapore.
Right; conquest and immigration are two different things. And *should* be fairly easy to distinguish, but some societies seem almost eager to fuzzify that border and just cross their fingers that immigration doesn't turn into that other thing.
Which is to say, I understand the appeal of Amelia rather more than I'd like to.
>but some societies seem almost eager to fuzzify that border
Notably the government of the UK - which seems intent on at the very least tolerating, and perhaps even encouraging, calls for Sharia to replace British law from Muslim immigrants.
I think Amelia's views go a bit overboard (though I'm writing from the USA, so I see assimilated immigrants, who have adopted American values and customs, as harmless - which differs a bit from the somewhat more insular views Amelia is an avatar of). Though any policy resulting in people like Jihad Al-Shamie being in the UK is a terrible mistake.
I think what strikes me as objectionable regarding that quote is that cultural competition probably isn't going to work very well if the participants don't agree on the rules, the limits of this competition, or even what constitutes winning, and if you put people of disparate enough cultures together, they're inevitably not going to agree on those things. For example, people of different cultures seem to have a wider range of views than might be hoped for on the question of when suicide bombing becomes a legitimate tactic to employ in pursuit of one's political goals. I guess I'm an enemy of freedom if I don't really want to engage in cultural competition with people who have significantly different views on that subject than I do?
Move over a bit, Gordon, I need to join you in the Corner of Shame Shame Shame 😁Ooh, do we get special hats to wear? I never got to wear a dunce cap before!
> If a culture of cooperation meets a culture of defection the defecting culture will “outcompete” the cooperating culture.
First of all, 'that's not how this works, that's not how any of this works.' Cultures are by definition insular groups interacting more with each other than with people outside the culture. Cultures don't compete by playing games with each other, so much as by playing games internally and competing on performance metrics. A 'defect culture' mostly means a lot of defectors interacting with each other, which will do worse than a cultures comprised of lots of cooperators interacting with each other.
Second of all, even if it *did* work the way you're modeling here, there's a reason defectbot doesn't win those competitions. The point of cultural competition isn't to have exactly 2 cultures and see which one wins, the point is to *discover tit-for-tat-bot (or whatever the actual optimal strategy is) through *lots* of cultures competing *and mutating*.
> A 'defect culture' mostly means a lot of defectors interacting with each other, which will do worse than a cultures comprised of lots of cooperators interacting with each other.
This is pretty much exactly what we see, no? Certain cultures lead to high degrees of educational attainment, others lead to female genital mutilation and oppression. Cultures that respect individual rights, and have a sense of collective responsibility beat out those without much respect for individual rights, and are mostly every man for himself (at least every clan for themselves).
> The point of cultural competition isn't to have exactly 2 cultures and see which one wins, the point is to *discover tit-for-tat-bot (or whatever the actual optimal strategy is) through *lots* of cultures competing *and mutating*.
A tit-for-tat-bot in this context would be a culture that cooperates, and punishes defectors. I.E. “If you don’t conform to the cultural norms that make us successful, you do not get to participate in our society.” Except that this sort of strategy is prohibited by open borders, since you *can’t* exclude people and cultures that “defect”.
The whole concept of cultural competition + open borders seems obviously wrong, since I don’t see any successful method of exclusion other than “here is a border and you can’t come in.” There can be no tit-for-tat strategy without the possibility of excluding (or harming) the defectors. So a cooperate strategy is a major loser in the long run.
...yes, it's just the opposite of what you predicted would happen:
>I’ve seen Bryan Caplan’s view on this stuff, and I’m not sure I 100% understand it, because at face value it seems obviously wrong. If a culture of cooperation meets a culture of defection the defecting culture will “outcompete” the cooperating culture.
>There can be no tit-for-tat strategy without the possibility of excluding (or harming) the defectors.
The issue being that you're inherently framing it in terms of collective punishment for a whole race/nation/etc, and conflating 'culture' with 'race/nation'.
Central to Caplan's idea here is individual punishment rather than collective, and the idea that any race/nation will comprise many different competing cultures, and any individual can decide to join any culture if they find it more appealing than their current one. That's how the competition leads to differing base rates over time.
> ...yes, it's just the opposite of what you predicted would happen:
Then you misunderstand me.
If you look at cultural competition as a process where self-contained cultures are competing with each other, with culture A being full of cooperators and culture B being full of defectors, the cooperators as an entity will do much better. That's not what I mean though. If you look at culture A and B existing within the same society, interacting and mixing, the defectors will make the cooperation strategy ineffective. It will be weighed down by free riders.
> The issue being that you're inherently framing it in terms of collective punishment for a whole race/nation/etc, and conflating 'culture' with 'race/nation'.
I don't think people should be judged collectively. I think they should be judged individually and allowed to immigrate given the intention and capacity to contribute positively to wherever they are immigrating, which are most people in this world.
I simply don't see it as viable for there to be cooperative cultures if there aren't fences around who can join and become a free rider. Caplan seems to frame the issue of immigration as one of freedom of movement, and not one of the strong economic incentive for most of the world to move to high-welfare societies.
If you exclude people from joining your political and welfare system, but have open borders, then you get a large population of people living side by side that are essentially separated castes. The gulf micro-states seem like a good example of this. If you don't exclude people from your political and welfare system in a society with open borders, the flood of free-riders (and their corresponding political power in a democracy) threatens the economic welfare and institutions that made it a desirable place to immigrate in the first place.
> In liberal democracies, with welfare, it seems impossible for the open-borders world Caplan advocates for to function. Population equates to political power, and there is little incentive to *not* game welfare when you have no preexisting social bonds to the society you’re receiving welfare from.
Doesn't this already happen with non-immigrants? Preventing immigrants from accessing these programs (which is already generally the case) seems like it would make this case *easier* to solve than the one where citizens abuse it.
In any event, sometimes culture just changes. Might not have to do with immigration or anything that can be reasonably prevented. Scott's written about this phenomenon before (https://slatestarcodex.com/2016/07/25/how-the-west-was-won/). Having institutions that are resilient or even anti-fragile to cultural change is far more important--and indeed, this is *kind of the whole point of individual freedom and liberal democracy.* Trying to preserve 1776 American culture in amber would be like teaching Newtonian mechanics instead of relativity in the name of science. Allowing culture to adapt is exactly how you overcome issues like losing out to a defect culture, not fighting an impossible battle to suppress opposing culture.
#15 - AI power generation problem - good discussion that “we” collectively will be fine but there’s such a race that the players won’t stop complaining because complaints get your project higher up in the queue
Re 55, I wouldn't necessarily be surprised if property crime had dropped much more then violent crime. I think of property crime as being much more incentive driven and faddish then say assault or murder which more emotionally driven on average. The ubiquity of cameras, security systems, surveillance, theft guards on cars/phones etc. make many types of property crimes riskier and less lucrative (assuming your in an area the police will pay any attention at all).
On the flip side, rather then a change in reporting rates for property crimes, it may just be compositional change from types of property crime more likely to be reported (break into a house / car for ex.) to ones less likely to be reported, or even detected, (shoplifting, porch pirates, that whole Amazon packages on the train thing).
e.g. people who used to break into houses are now porch pirates and shoplifters as the former will get you jail and the later will get you, at worst, a disapproving glance.
#4: It reflects a substantial decline in lab leak theory among the sort of extreme weirds who think prediction markets are the coolest thing ever. And sometimes those weirdos have their predictions anchored somewhat in reality (or in other people's perceptions of reality) by the prospect of making or losing money. But as you point out, it's extremely unlikely that this claim is going to make anyone any money any time soon. So there's no more reason for their predictions to match reality than e.g. a fantasy football league's. And like the typical FFL, the real incentive is to feel like you belong in the tribe.
That lab-leak is declining among the rationalist-adjacent weirds who do the prediction-market thing, is I think driven by the Rootclaim debate and the discussion surrounding it in rationalist circles (like here). We can disagree as to how much people should update on that debate, but I'm not seeing this finding adding anything new.
Are the “suburban neighborhoods on top of giant multi-story buildings” pictures AI slop? There are cars up there. How did the cars get up there? Why? How will they get down?
Cars get to the top floor of parking garages by way of ramps that take up a large fraction of the internal volume of said garages, and which incorporate still more parking space. That's a huge cost and architectural constraint to impose on a large apartment or office building that really wants to have its internal space configured very very differently.
And to what purpose? Yes, the people in the "suburban neighborhood" want to have cars. So do the people in the apartments/offices, so there's ultimately going to have to be parking for all those cars. Probably underground. But in any event, there's no reason for the suburbanites to not park their cars in that same place, and it would cost them hugely to indulge any "but *my* car has to be in *my* garage or it isn't *really* suburbia" fetish.
I would be very surprised if anyone building a quasi-suburban low-density housing project on top of a large apartment or office building, were to incorporate streets and driveways and garages. I would be not even slightly surprised if anyone trying to bash together a quick visualization of a "suburb on top of a highrise" were to lazily copy an existing suburb or suburbs.
There’s a style of apartment building called a “Texas donut” where the building occupies a full block, there are apartments around all the edges, and a parking garage in the middle with whatever sort of ramp goes up it. This is popular because it means that everyone can have their parking space right next to their apartment, and it makes more efficient use of the land area by using the interior part of the block for parking rather than needing to cut out bits of the building envelope to ensure there are windows in the center.
If you just take one of these Texas donuts, and put some townhouses on top, it looks like this.
Also, I think "ramps are also parking spaces" might just be an American thing? I don't think I've ever seen it in Europe. Not allowing parking on the ramps would make them a lot smaller.
> it would cost them hugely to indulge any "but *my* car has to be in *my* garage or it isn't *really* suburbia" fetish
Sure, but these are by local standards ridiculously expensive houses for rich (but not ultra-rich) people, it's okay that it costs them dearly. They are okay with paying extra for the convenience of being able to park right outside their house.
I don't understand why the inconvenience of spending several minutes driving up/down ramps (especially during busy periods) outweighs the inconvenience of walking a short distance to and from an elevator. Or does the building not have enough elevators?
Because you have to do it carrying stuff, or children, or freaking Ikea furniture.
Also there seems to be a special ramp straight from ground level to the roof.
I don't get why people are questioning this. Have they never lived in a place where they don't have their car parked right outside their front door, or have they never lived in a place where they do? Because I gotta tell you that was one of the best things about upgrading from an apartment to a house.
I'm just trying to think about the daily tradeoff. A dedicated 10 story ramp would make sense if elevators are not available or far away, or if power is frequently unreliable. The article seems to focus on avoiding street level flooding as a bonus but in a flood one would be stuck and unable to leave and power likely out. The whole thing sounds dystopian.
Alternatively, they simply took an existing neighborhood, jacked it up, and constructed a big building underneath. The cars are owned by the sort of people who don't read paper notices attached to their doorknobs.
13. The $170 billion number is just for "23 of the world's poorest countries with high-quality survey data." The estimate is $318 billion per year to end poverty worldwide.
New estimates suggest that lifting everyone above the $2.15/day extreme poverty line would cost about $170 billion per year, or 0.16% of global GDP, in the 23 poorest countries studied, which collectively account for around half of the world’s extreme poor.
The analysis then uses the results from these 23 countries, all of which have recent and reliable household survey data, to estimate the cost of ending poverty in countries where such data do not exist or are not publicly available. This extrapolation implies that it would cost $318 billion per year to reduce the global poverty rate to 1%. This amount corresponds to 0.3% of global GDP.
Re #39, about traditional societies sucking pretty bad. For those of us who got a rosy view of cultural evolution via The Secret or our Success, have a good skim of Sick Societies by Robert Edgerton. Traditional suckage is, unfortunately, pretty much the default.
Thanks. I saw that, and I'll grant it's possible you're right about the literal lizardman example (or partly right), but I think my overall point still stands and there are examples that are even less defensible than the lizardmen. For example, the 5% of Obama voters who said they thought Obama was the Antichrist, or the mischievous responders in https://pmc.ncbi.nlm.nih.gov/articles/PMC6215371/ . I do think your point that some polls adjust for this carefully is well-taken, and that we may be able to take those polls in particular at face value (although I don't know enough about this to be sure, or to know which polls these are).
Thanks for taking a look! I certainly don't claim that there are never survey results we should dismiss, but I do think the Lizardman Constant can be an impediment to maximizing what we can learn from polling.
When I look at a list like this from Randal Monroe I see a lot of 5% or less results that seem very reasonable to me (surely a nontrivial percentage of people are fine with phones in movie theaters, dislike the olympics, or are less than satisfied with their friends for instance). But several smart ACX fans in the comments are so wedded to the lizardman concept that they are unable to appreciate this.
> the 5% of Obama voters who said they thought Obama was the Antichrist
But in serious theology, "the Antichrist" is not a supernatural hell-spawn trying to take over the world and bring about Armageddon — it refers to any worldly peace-bringer whom the masses come to worship with the kind of fervor they should reserve for the Messiah. It's not inconsistent to vote for a politician while believing him to be "the Antichrist" in this sense, so long as you're voting for him in a pragmatic way without getting taken in by the pseudo-religious hype. I can easily imagine a Christian who votes for Obama due to viewing his platform as the lesser of two evils, but is very concerned at the hero-worship of him and how it detracts from Christian humility, and expresses this as "Obama is an Antichrist".
You may be right! I was also thinking that over the phone it is easy to mishear or misunderstand “the antichrist” as being anti- (or opposed) to Christ. Particularly if some respondents are not very knowledgeable. This is not deliberate tampering with the survey or even really an insincere belief, just a related concept that in an ideal world would be easier to separate out from what they were going for with that particular question. This kind of mistake doesn’t necessarily suggest problems with other questions on the same survey either.
I noticed the phrase "champing at the bit" showing up much more recently than ever before. Is it one of those things where you learn something new and suddenly it shows up more often because you notice it, or has Scott really been using it more? Or does the AI progress give many more opportunities to use it?
Probably the first. Google ngrams says that both "chomping" and "champing" at the bit have been trending down slightly for the last eight years. Most sources correctly used "champing" until 1999, when "chomping" overtook it in a rise that had started in 1965 or so. "Chomping" is currently about 70% more prevalent than "champing". Alas.
I actually hadn't noticed Scott using either one, but I also didn't just learn the difference. :-)
60: Some extra speculative context on cancer and Alzheimer's:
Trade-offs between cancer and degenerative disease crop up fairly often. Both of them are different manifestations of increased cheater cell load in old age.
Young bodies are very good at repairing the body through regeneration. Old bodies tend to have very weak regeneration. There are some reasons to speculate this could be adaptive - weaker old-age regeneration resulting in longer overall lifespans.
Each cell in a multicellular body has an ability to bring down the entire organism by hotwiring its "multiply right now" button to always be switched on. As a result, evolution of multicellularity resulted in all sorts of permission-slip style systems that strictly control when replication is allowed: tumor suppressors.
Young bodies have the same genome and the full set of tumor suppressors in all cells. It's a high-trust society where the risk of having a cheater cell is low. If tissue gets damaged and some cells get lost, it's pretty straightforward to tell surviving cells to proliferate to close down the gap, and expect that they will proliferate in a responsible orderly manner.
Old bodies are riddled with mutated precancerous cells - cells that have some of the locks removed. The term for this state is somatic mosaicism. Somatic evolution pushes precancerous cells to out-compete and out-multiply normal cells, so over time their number is constantly growing. Mosaicism is pervasive in proliferating tissues like skin and intestines, but is also detectable in largely post-mitotic organs like the brain.
So evolution encounters a trade-off between "being healthy" and "surviving longer":
1) The old body could give cells the permission to proliferate and repair the tissue. This keeps the body functional, but at the same time it's removing locks that prevent the precancerous cells from going "all in" and destroying the body by forming tumors.
2) The old body could enter the "coup-proof" curfew state, and prevent all cells from regenerating by triggering cell senescence or tissue inflammation. This "low trust" tissue state protects from cancer, but leads to the frailty and tissue loss common in many diseases of old age.
To massively oversimplify, if you die of cancer at 60, your body chose option 1. If you die of degenerative disease at 80, your body chose option 2. If you live to 110, you were somewhere in the middle, plus were lucky enough to avoid a cancerous cell getting all of its tumor suppressor locks mutated.
This "aging as a tumor suppressor" idea is rearing its head in biogerontology from time to time, usually when the new crop of anti-aging therapeutics fails to deliver results. To be clear, it's far from a consensus view - but to be fair, there's almost nothing in biogerontology that's consensus.
One question is "Why would this regeneration-cancer trade-off apply to Alzheimer's when neurons don't proliferate anyway?" - I'm not sure. I'm not very familiar with brain aging, perhaps someone could expand here. Under "low-trust conditions" lens, brain degeneration could be an unfortunate byproduct of the pro-inflammation endocrine factors clamping down on somatic mosaicism in rapidly proliferating tissues like skin, hematopoietic system, or digestive system. More relevantly to Cystatin C, remodeling of extracellular matrix is another surface for a trade-off: making ECM more rigid in old age helps encapsulate nascent tumors, but also solidifies all sorts of plaques. So you'll see cancers expressing pro-metastasis factors that appear to have "anti-plaque" function. That's like having battle engineers of the invading military force repair bridges that were blown up by the country's own defenders.
Cool concept — it'd be interesting to see an exogenous DNA repair mechanisms that detects and repairs somatic mutations. Or a method to detect less-mutated cells that are in proximity to heavily-mutated cells, then simultaneously inducing apoptosis in the mutated cells and inducing the less-mutated cells to fill in the gaps. (at least, less-mutated along tumorous axes)
The difficult thing about the first approach, assuming you figure out delivery, is figuring out how to not create more mutations than you fix. The tools we have for DNA modification are prone to error, even more so in the uncontrolled environment of a living body, and that problem gets way worse when you're trying to edit oncogenes. 0.01% chance of error * 100,000 cells = whoops, you might've given the patient cancer.
The difficult thing about the second approach, assuming you figure out how to identify heavily-mutated cells and deliver your medicine, is that precancerous cells commonly have mutations that resist apoptosis. That's one of the major functions of cancer-related mutations. If you apply some apoptosis-causing agent + some growth promoter to a population of cells, a small percent of precancerous cells will evade apoptosis and end up exposed to the promoter, and whoops you gave the patient cancer again.
We might see some cool and innovative cancer prevention measures like this eventually. Just emphasizing why this research is so hard.
Would there really be any evolutionary pressure to keep the body alive longer in old age at the cost of increased frailty? I'd have thought in an ancestral environment you're likely to die anyway if you're making either choice, and in any case you're very likely not reproducing anymore. So my naive expectation would be that evolution just programs whatever works for younger bodies that it actually "sees clearly" and it "doesn't think about" what happens to older bodies very much.
You’re effectively describing Medawar's "Selection shadow", and generally, yes, you’re right, evolution won't design an "aging program" specifically for the post-reproductive life stage. But to see cancer-frailty tradeoffs, no new "programs" need to start in old age. The tradeoff can simply be the consequence of tumor suppression strategies that work the same pre-reproduction as they do post-menopause.
What's happening here isn't evolution looking at the 80-year-old human and deciding "Let's make him frail to keep him alive until 90". What probably happened was evolution looking at the 10-year-old proto-mammals who keep dying of carcinoma and deciding "Let's give these creatures a hair-trigger tumor suppressor so that they survive until 20 more often". You can expect lifespans to rise in this way as species shift to ecological niches that reward slow life strategies: memory formation, sociality, caring for the young, predator avoidance, niche construction, basically anything where benefits to your kin compound over organismal time.
To achieve that, evolution selects for aggressive tumor suppression in response to cheater cell burden. Long-lived animals evolved to shut down replication at the slightest signal of DNA mutation. This adaptation is highly useful for a pre-reproductive organism and lets the young survive the cheater cell burden typical for their age. The "trade-off" is just that this same hyper-vigilant safety lockdowns keep happening more and more often as the number of mutated cheater cells rises steadily over the decades.
So under this lens, evolution didn't "program frailty" for the elderly. It programmed "high-security anti-cancer lockdowns" for the young, with side effects that ramp up over time until they're just as deadly as cancer itself.
(As an addition, there are some good reasons to believe post-reproductive organisms can also be visible to evolution via kin selection - see grandmother hypothesis - but this is not strictly necessary for the explanation)
Re: #55 - Does anybody have a link to a steelman on the claim the murder/general crime rate actually has not gone down? I'm interesting in seeing a good version of this argument - the simple version (gunshot wound care is so good now it's just hard to kill people) falls flat on its face for the reasons Scott mentioned in the post. It also defies casual observation for me - you'd never hear about people going to much of Oakland and neighborhoods like Third Ward, Houston that didn't live there when I was growing up and now they seem mostly fine? And despite living in a kind of crappy part of Houston, not exactly a low crime city, for many years I know don't know a single person who has even gotten so much as held up for their wallet? N=1 of course, but claims about omnipresent danger in American society are so common online it makes me wonder if I'm not just weirdly lucky.
I'd like to see that particular steelman too. Wouldn't have to be all or nothing e.g. a solid steelman for current reductions just being seriously exaggerated, would be worth having.
The reported/gathered drops in US crime rates starting with 2023 have been _so_ drastic -- and so broad, and showing no signs of stopping -- that it's hard _not_ to wonder whether there isn't some big data confounder that's being overlooked. I say that even though as a lifelong big-city dweller my personal eye test is similar to yours.
For example Jeff Asher's stuff on Substack, which appears to be high-quality and thoughtfully done, is lately making my brain go "naaaah, couldn't be. You gotta be missing something."
Holding people up for their wallet is much less useful as a crime these days! A lot of people carry little or no cash, and credit cards and phones are easier to disable remotely than they used to be.
Meanwhile, online fraud is much easier and more lucrative than it used to be, particularly if you do just a little bit more research into how crypto works than the average gullible person.
Certainly, and that may be a driver of why we see less petty muggings in the data - the value proposition just isn't there anymore!
As far as your (potentially not totally serious?) suggestion that these people may have gone into crypto scams - I can tell you grew up in a nice neighborhood haha. The kind of people who'd jack my stuff walking home from school as a kid are not going to understand how to rugpull people in a crypto scam. *Maybe* some simpler online scams, but I'd also question how much those really happen. Even back when the US was much poorer, you had to be truly dumb to have to rely on petty crime to get by. Criminals smart enough to figure out how to spoof an email and convincingly clone an e-commercd sites
UI were never the guys sticking a knife in your face for a wallet in some counterfactual where the Internet doesn't exist. They'd be involved in some kind of white collar crime or organized gang that consciously avoids doing things like sticking knives in people's faces, especially completely random people.
Good points! My main point was just that there are some types of crime that have likely gotten easier as ways to make money, even as others have gotten harder. I don’t know which ones would have real substitutions effects. But I wasn’t thinking of a crypto rug pull where you sell high after getting others in - just the kind of crime where you get someone to send you their wallet info or wire you some coins by pretending to be a Russian bride or pretending to sell them some drugs.
Oh sure, that makes sense. I wasn't thinking of those. Yes, those kinds of crimes probably substitute much better in terms of simplicity, risk, and reward (although from a social outcome POV, I'd much rather street criminals scam people via lying in emails than actively threaten physical harm)
For #32, CBS claimed it wasn't suppressed but had enough fact checking issues that they needed to postpone it for verification (and they were always going to release it eventually).
Ymmv on how trustworthy you find CBS, but from the outside view - if you assume an honest non-politically motivated newscaster management exists (even an intermittently honest one), you should expect to see stories like this happening (since individual reporters who lean left will occasionally (a) make a lot of trump-critical reports with twisted facts and (b) claim to be suppressed in the face of fair pushback).
I’m no expert on clinical trials but I wonder whether there’s a dynamic somewhat similar to what I see in environmental reviews of infrastructure projects. A lot of money is at stake and project sponsors are extremely risk-averse, so the analysis often goes far beyond what is actually required by the applicable regulations in an effort to make the process bulletproof against lawsuits (even though NIMBYs often sue anyway just to delay and make the project more expensive).
I am an expert on clinical trials. This is one of the dynamics at play, but not the only one. I'm planning to write a more thorough response under the linked article. It has some good insights, but also some of the things he mentions are not right at all.
I spent a couple years programming software for data collection to support clinical trials. Running on Palm Pilots - to give you an idea of how long ago. What the linked article doesn't mention is that electronic data collection is much better than paper-based because it avoids what they call the "parking lot problem". The subjects fill in the last months worth of questions just before they turn in the papers rather than when they were supposed to fill them out - the electronic method forces them to answer the questions in a timely manner. I would have expected this superior methodology to win out over time, but now I'm not sure.
Ah yes, "parking lot compliance". I've had coordinators tell me how they sometimes watch patients sitting in their car filling out the patient diary.
Nice thing about ediaries is that they remind the patient when they should fill them out. Problem is older patients don't understand them.
The diary data is usually an exploratory endpoint (since it's unreliable), so nobody wants to spend the extra money on them. Also paper seems like it should be cheaper (before the extra SDV time). Sometimes we pay for them if they're important to the primary endpoint, but usually they're not so it's a hard sell to spend extra money on a vendor in order to save money on data entry and verification.
> Oh well, I’m sure tech journalists are right about everything else being a giant Ponzi scheme that will inevitably collapse and reveal the entire tech industry to be an emperor without clothes.
You picked a hell of a day to say this.
(Context: Bitcoin crashing at an insane rate this morning, to the point people are talking about the crypto bubble popping)
I am actually happy to see more bans of low-effort high-temperature comments.
I agree that it's a bit atypical, but Scott also engages more with politicized discussion in this post than usual. And I don't think the bans are outside the usual policies, it's probably just that Scott happens to have more time/energy/engagement today.
AI is a scam? Dude, it's your problem that you can't use it. For me, it generates websites and translates books, at the cost of €20 a month. But if you think it is a scam, simply ignore it. I know people who refused learning to use computers 30 years ago, because they believed that it was a fad that would soon pass.
(And the Bitcoin "crashing at an insane rate" is still 20x as expensive as when I bought it, and I expect it to stay that way.)
If this debate is supposed to be "rationality-adjacent", at least the participants should not clearly contradict reality and be smug about it.
Cursory search suggests this might be relevant to #55. In-hospital mortality from GSW has fallen, but pre-hospital mortaility has increased so in net GSW mortality hasn't fallen much.
Not very consequential, but funny: The super-secret Coca Cola formula has apparently been duplicated, with some mass spectrometry and a lot of testing. Since it was never patented to avoid publishing it, anyone can now make something indistinguishable from Coke. You don't even need coca leaves. https://www.youtube.com/watch?v=TDkH3EbWTYc
Regarding 21 apparently, we already have a timecapsuledLlm only trained on data between 1818 75, although it’s apparently a small model as you might expect. Unfortunately, I can’t find any copies of it online, which are directly usable, although that’s apparently in part because I’m not a programmer since Grok assures me that a normal programmer can easily use the publicly available data to run a copy of this. Still, it does seem like that kind of thing you can use for interesting experiments and trying to find out how people would have justified their world view back in the 1870s, which does strike me as super interesting. Actually, if anyone has a clue, how one might go about turning what is publicly available into something I can prompt. I would be extremely grateful because right now grok’s explanations while pretty confident leave with even more questions as it keeps using and talking about concepts, I have never heard of. Maybe I should try asking ChatGPT or Claude instead.
I don't think your link supports your point - it says most forecasts cluster 20 years out, but now the forecasts are 2 - 10 years out, presumably because AGI is actually getting sooner. Over the past few years, most people's timelines have gotten shorter rather than longer - see https://agi.goodheartlabs.com/ .
> #22, New representation-in-historical-movies controversy, this time about an African woman getting cast as Helen of Troy in the new Odyssey...
Funnier still: my friends who are expat Greeks are now asking the question of why many of the major actors don't look Greek or of Greek extraction, with Matt Damon being the most obvious Vóreioi in the film.
I think it's fair to say that there's a spectrum of Greek-passingness, with Jason Mantzoukas at one end, then Omar Sharif, then Orlando Bloom, then Matt Damon, then Zendaya Coleman, then Lucy Liu, and then finally Nupita Nyongo who is just about the least Greek-passing person on Earth.
I've never seen the movie, but I looked up some pictures. He doesn't look totally implausible, but his face looks more Greek than his hair. Somewhere on a par with Sharif.
It certainly bugs me when historical dramas or dramas set in supposedly historical realities use actors with obviously non-period-authentic ethnicities in their cast. I found Bridgerton particularly discordant, to the point that I couldn't watch it. Primarily because it whitewashes the actual existence of slavery in that period. I don't have a problem with mixing ethnicities and sexes in Shakespearean plays, because even his history plays are mostly history-free. As for Homer, it's mythical enough for me to be to *mostly* unbothered by the casting, but it does seem that there should have been some ethnic consistency within tribes/kingdoms as Deiseach said below. Lupita Nyong'o would've made a fantastic goddess, IMHO—Athena or Circe.
I think Lupita Nyong'o would be fine in a part like, say, Calypso but not for Helen of Troy. Unless you're going to cast *all* the Spartans/Laconians as black Africans, which would include Menelaus and Agamemnon, which I could live with as at least it would be consistent and a different, interesting choice about how these are all representatives of separate Greek kingdoms and not some unified state called "Greece" that did not exist in that time:
That would be Diverse Casting that works! Different kingdoms are internally consistent but the Spartans don't look like the Macedonians don't look like the Thracians and so on.
Plus, Nolan's costume and set designs seem to be deliberately trying to make everyone look as unattractive as possible, so we're in the "mud and earth tones and black leather" past of TV and movies, and not the "people liked colour and gold ornaments" past.
That would be an interesting casting challenge. Out of curiosity, I did a cursory search to see if there was a substantial genetic change in modern Greek populations from their Bronze Age predecessors, and Greece is one of those rare places in Europe where there is substantial genetic persistence. Usually, conquerors and migrations shuffle up the genetic deimcks, but, depending on the region, populations are 70 to 90 percent similar to their ancestors (at least that's the claim made in this popular article). And overall Greeks from the islands are genetically similar to mainland Greeks.
>36: Indonesia has solved the conflict between density and single-family zoning by putting suburban neighborhoods on top of giant multi-story buildings (h/t @xathrya):
How many Indonesian kids need to die chasing their soccer ball off the top of a skyscraper before we start to care?
Also, who would call the depicted neighborhoods "suburban"? I live in a pretty dense suburb on the SF Peninsula, in a lot that is smaller than any of my neighbors', and I have more of a yard than that.
Heh. But seriously, didn’t it look more like the cheek-by-jowl townhouses of a city than like anything you’d call suburbia? Doesn’t mean it’s unlivable, but it’s a very different niche.
Surely the relevance of no 13 is that we are spending that much _every two years_ and have not meaningfully reduced absolute poverty in the countries we are spending the bulk of that money in?
But it is a bit of a vicious (or only partly virtuous) circle in that currently our aid money only makes a minor impact on demand for Harrods' shopping, custom made suits, expensive cars and watches, expensive hotels and restaurants in the local city center, and private jets. Which, apart from the hotels and restaurants, only modestly impacts inflation in these markets and only outside of the countries we are giving aid to.
Direct cash transfers on that scale in such dysfunctional economies would likely trigger substantial (NGDP growth and therefore) inflation in the local economies, which would raise the absolute poverty bar since people could no longer afford to subsist on the original amount.
But course we are spending this money _every two years_, and direct transfers would substantially jncrease demand for good governance, this would absolutely work to massively reduce absolute poverty even despite a few localised bloodbaths.
> Direct cash transfers on that scale in such dysfunctional economies would likely trigger substantial (NGDP growth and therefore) inflation in the local economies
That would only be true if they can’t get imports. If they can get imports, then the cash transfers end up turning into imports (though there might be some inflation in the cost of locally-supplied services).
That's the point though. Leaving aside that a _lot_ of the demand would be for construction and personal services, indofar as goods are concerned these countries can't quickly get any large amount of imports, their economies are not functional enough to facilitate that.
This could ofc be alleviated by ramping up the transfers over a few years but even then I predict substantial inflation.
51: When I wrote my post on using light glasses to treat non-24 (and that they can treat SAD), I specifically looked for any mentions in the ratsphere about them, and found none. This is confusing, because they are actually a viable alternative to lumenators! They cost the same or less, are portable, and don’t change in effectiveness based on gaze direction. They’re strong enough to entrain my circadian rhythm, and they have similar melanopic lux to lumenators. I want more people try them out and report on them!
Hey Drossophilia. I'm also on team non-24, but I've gone into 'remission' (if that's the right term) as an accidental side effect of taking a low dose of Escitalopram. It was weird, in bed every night at ten. That lasted about six months before slipping back into delayed sleep phase, where it has stayed locked for a few years so far. I haven't found any studies that mention this effect, but I thought I'd mention it to you. Cheers
Oh wow! I have a somewhat similar effect taking bupropion and memantine (is what I think is happening). Unfortunately it makes me only sleep for 6 hours, so I’m quitting the latter. Thank you for sharing your anecdote!
The theory I have right now is that non 24 is like ADHD for sleep. One of the major side effects of Escitalopram is apparently 'insomnia' during the first few months. So maybe just as stimulants are a good treatment for ADHD, perhaps 'insomiagenics' (to coin a term) might be effective treatments for non 24? No idea really.
I haven't studied Soviet history, but just going off Sarah Paine's lectures here :) I assume we have to keep the extractive nature of the whole enterprise in mind, or else the numbers feel backwards. Like if we lost "Calikraine" with 18% of our population and 16.2% of our GDP, sure that would be a loss in terms of total national power and prestige, but wouldn't our relative GDP-per-capita go *up*? Well maybe in Loose-US-Analogy-istan it would, but not if Moscow (Moshington DC? New Mork?) had been systematically siphoning wealth from the provinces.
Relative GDP per capita is a thing, but it's not the only thing. There's also absolute GDP, and there are some people who care about absolute GDP more than they care about GDP per capita. That's a particularly common mindset who expect to be at the top of the pyramid, with all of that GDP being channeled through them.
If you had a choice between being the founder/owner of a billion-dollar-a-year corporation with ten thousand employees (GCP per capita: $100,000) or a million-dollar-a-year startup with five people (GCP per capita: $200.,000), which would you choose? Which would you expect most other people to choose?
Fascinating. Question for the mathematicians out there with respect to #4. Suppose we have an ideal betting market wager which can only resolve in one direction within a finite amount of time (e.g. this can only be proven true, but if it's false, it will remain open forever). What sorts of dynamics would we expect from such a wager, absent any new evidence? (My intuition is we would expect something like exponential decay towards 0% as we approach the limit of infinite time).
If decay is the expected behavior, what interpretations can we make about the wager given the shape of the curve? What would determine the underlying decay rate? Would we expect the decay rate to change given new evidence?
This is a side of better markets I've never thought about before.
We cant make assumptions about the what the shape of the curve should look like without making assumptions about how people's beliefs update in absense of new evidence. There probably is be an assumption of at least one side that by xyz date, evidence should have come out, and if it doesn't then that is evidence going the other way, but we dont know magnitude or direction here.
I think its pretty dumb to have infinite horizon predicition markets, they should just do long horizon instead.
> There probably is be an assumption of at least one side that by xyz date, evidence should have come out, and if it doesn't then that is evidence going the other way
What if the question is about the Second Coming of Christ?
I wasn't; it's just an obvious counterexample to the idea that there should be "an assumption of at least one side that by xyz date, evidence should have come out".
I am interested in the article, but there's something I'm not following. The time crunch theory is described like so:
> right now, there's not much interesting stuff happening on Polymarket. People are spending a lot of money betting on sports, but not much else. But at some point in 2025, other markets will get a lot of attention. [...] Right now, all these markets have mere single-digit millions of dollars in trading volume, but that could very easily change.
> And if it changes, some of the people betting No on Christ's return will want to unlock that money -- that is, sell their "No" shares -- so that they can use it to bet on other markets. If enough people want to sell their "No" shares, the "Yes" holders may be able to sell out at an elevated price, like 6%, potentially getting a 2x return on their investment!
I'm willing to stipulate the following:
1. Christ will not return in 2025;
2. Everyone is aware of this;
3. The market will resolve NO on January 1, 2026.
I agree that the Second Coming market is a terrible position to hold because it gives you low returns over an unreasonably long investment period.
And I also agree that at some point some NO bettors are going to realize that and sell their NOs, and this will drive down the price of NO.
But the quote I just pulled says that this will somehow mechanically raise the price bid for YES. I don't see why that should be the case. The hypothesis is that a lot of people realize they have more NO than they want, and sell it. This will mechanically lead to a lot of other people buying NOs at lower prices.
But there's no rule that says that when NO is trading at $0.61, YES has to be trading at $0.39. The market could reflect a 61% chance of no Second Coming and a simultaneous 2% chance of at least one Second Coming. The fact that that totals only 63% isn't a problem; we've already stipulated that the true probabilities are 0% YES and 100% NO, and 2% YES / 61% NO isn't epistemically worse than 39% YES / 61% NO - it's epistemically better!
Where would the increased demand for YES come from, in this theory?
Fair point. I'm pretty sure that in practice the probabilities always sum to 100 - there's always someone on the other end of the bet - but I don't think I understand Polymarket well enough to properly grasp why that is.
The fundamental working of Polymarket is that you can pay $1 (measured in a USD-tied stablecoin) to receive a matched pair of YES/NO tokens.
In "ordinary use", one of those tokens will be redeemable for $1 (of stablecoin) when the market resolves, and the other one won't.
Turns out Polymarket will allow you to redeem a matched pair at any time, whether the market has resolved or not. Which means that if you're strapped for cash and want to sell your 100%-chance-of-NO for $0.94, Polymarket's system can match you against someone who wants to buy that NO for $0.94...
...but it can also match you against someone who wants to sell a YES for $0.06. In that case, you buy their YES for $0.06 and redeem the YES/NO pair for $1, netting $0.94. This is why depressing the price of NO automatically raises the bid for YES.
35: the "mental health" vs "mood" thing may have explanations that aren't merely stigma per se. By default if I saw these on a questionnaire I would see them very differently.
The brand new 2025 Honda Civic you bought is functioning perfectly, but it does not feel as good to drive the Civic as it would feel to drive a brand new Corvette. There is nothing wrong with the health of that Civic, a mechanic can do nothing to make it feel like a Corvette, the Civic is high health + low mood. Maybe liberals' problems tend to be more defects that are fixable, or maybe ~10% of them tend to think everything is supposed to be awesome and perfection is attainable and why aren't we constantly working to make everything sunshine and rainbows for everybody and if it's not sunshine and rainbows for me then there's a defect somewhere. And maybe that ~10% of conservatives are just more likely to accept that sometimes live hands you a Civic instead of a Corvette, it's not gonna magically become a Corvette, none of us like the Civic but it runs as intended, that's life so get in the Civic go to work and shut up.
36: I feel like I've read discussions about putting neighborhoods on top of dense buildings and there's some practical reason you aren't supposed to do this. Sewer lines or water pressure, fire hazard, windbreaks, carbon monoxide from heaters below you, something about heat radiating off the surface, I don't know but intuitively it seems like there's some obvious flaw in this plan.
Re: #12, I think a string use case for AI is answering real life halachic questions on demand, replacing texting a rabbi or searching through long Q&As for something similar to your question. I propose to create an AI benchmark for that. And it shall be called... BenchBench. :)
I think 'women asking men out' has a strong cultural valence too. When I (a New England boy) went to college in California, I was astonished to find California girls coming up to me and asking me out. Totally unprecedented, unheard-of behavior in much more traditional New England (and I think the rest of the country too). I had no idea how to handle it either- often I wasn't interested but it was awkward and difficult to let them down gently
The historic German Reich (I.e. the Holy Roman Empire) included Bohemia and Moravia. The cities of Prague and Brno had historically been majority German speaking until the mid nineteenth century and were still regarded as centers of German culture even as late as the 1930s (even if a lot of that “German” language culture was being produced by Czech Jews). Notably, Germany never tried to incorporate Slovakia, which also had German minorities but had historically been Hungary, not part of the Reich. Maybe that had some effect on Hitler’s thinking.
#13: I think this just fails a very basic intuitive test: the flipside of poverty as a state of low levels of consumption is low levels of production, so trying to fix the consumption part via mass fiscal transfers without having any plan to fix the production side is not a recipe for "ending" anything at all. At least Jeffrey Sachs had a theory that addressed these pretty basic elements, even if he turned out to have mostly been wrong.
#27: those are all very nice, but there's a suburban mid-rise apartment complex going up by me with an exterior that mixes four different colors, so I'm certainly not sweating it, India.
#32: I don't have any inside info here, but I think it's likely that your characterization is inaccurate and that this was really an attempt by Blob journalists at CBS to discredit their non-Blob boss, Bari Weiss, for the crime of interfering with Blob output in a formerly Blob-owned factory.
Re 13: lack of capital is one of the major limits on production. If you can buy some sheet metal to set up sturdier walls and roof, then you can spend less of your time doing repairs and more doing services or labor for others. If you can get effective treatment for illnesses then you spend less time out of education and work for sickness. If you get a butter churn or sewing supplies, you can spend a bit of labor to upgrade the raw goods you produce so that they are more valuable for others and sell them for more profit.
Agreed, which is why "just give everyone money" seems unlikely to work. You don't know what portion the recipients are likely to spend on capital goods. The fact that these people are living in extreme poverty in the first place suggests they probably don't have access to many capital goods at any price. Where they live, you can't just go down to the farm supply store and buy a butter churn; that store doesn't exist within any reasonable distance.
Huh, I would have thought the opposite! If you send a lot of dollars to rural villages, then people in developing cities now have an incentive to set up a farm supply store in the rural area, as a way of getting dollars. This lets the market create the supply chains that actually bring poor people into the world of capital.
It's possible that higher coverage induces more aggressive litigation and so claims experience is not independent of coverage. But that's a separate case to make.
(Also, I've been told by others that Empower's driver base is essentially just "drivers who got kicked off of Uber and Lyft", which, if true, makes me pretty sympathetic to DC here.)
If you call one of the big personal injury firms, they’re gonna route you through a pre-screening where a newly-minted lawyer will make a quick estimate of what he thinks your claim is worth. Part of that math involves assumptions about typical coverages, and obviously deep pockets increases the estimates, if you think the at-fault driver was in the course of employment at the time this would raise it of course. I imagine if industry standard went up 2x you’d find all the component parts of a claim drift up to fill the difference, much as US college tuition rose to whatever level of government loans were offered.
#36: During World War II Boeing built an entire (fake) 26-acre suburban neighborhood on top of its Seattle bomber plant to conceal it if there were Japanese air raids. The “houses” were only about six feet tall, as bomber crew high above would not be able to tell the difference.
> if trauma care has gotten twice as good, then the apparent number of murders will halve even for the same amount of crime
Does this not assume that all murder victims somehow end up in an ICU? I would assume that in the vast majority of murders, the victim is already dead before anybody knows about it, much less is able to intervene.
If only 20% of murder victims actually end up in the hospital at all, then a 50% reduction in attempted murder victims whom trauma care fails to save becomes only a 10% reduction in murders. Maybe my intuitions are wrong here and more murder victims than I expected end up receiving hospital care before they die.
Yeah, it seems like for both murders and accident fatalities, improved medical care/improved EMS has to be decreasing the numbers over time, but probably not by a huge jump at any one time.
One interesting thing to do would be to look for some natural experiments here. Like, when an ER opens/closes in the area, do we see the result in the local murder rate/traffic fatality rate?
The source for the $40 billion savings number from the eye medication study says:
> The use of bevacizumab from 2008 to 2015 resulted in an estimated savings of $17.3 billion, which corresponded to a $13.8 billion savings to Medicare and a $3.5 billion savings to patients. This amount underestimated the actual cost savings to Medicare providers, since approximately 30% of Medicare-eligible recipients received care within Medicare Advantage plans and were not included in this analysis. ... Additional savings over the $17.3 billion would have accrued from the use of bevacizumab if diagnostic categories such as diabetic macular edema and retinal vein occlusion were included in this study.
The number for the "additional savings" is quoted elsewhere as in excess of $30 billion, presumably that got rounded up to $40 billion. But importantly this is savings over 7 years, so keep that in mind when comparing to the *annual* budget of the NIH, and that a good chunk of the savings is for the patients not Medicare.
I want to thank Amelia for briefly turning the Culture Wars from something relentless and insufferable into something f@#%ing badass. Just look at this thing: https://m.youtube.com/watch?v=3B4524ot5BM
It's so sad. If the Amelia thing had happened a few years ago, some gifted right-wing troll would have created amazing videos out of it, perhaps by doing Amelia-themed alt-right karaoke of recognizable songs, as Walt Bismarck used to do with Disney songs back in the day.
Whereas all we get now is slop.
Slop that, granted, one day will start to feel truly human, and that will be an even sadder day, because it will mean humanity has been replaced.
The alt-right has it wrong. It's not immigrants that are stealing our jobs. The real great replacement, the real genocide, is the one robots are doing to us.
Thanks very much for the shout-out, your writing on kabbalah is some of my very favorite ever of all time ever so it is particularly tickling to know you enjoyed (and hated?) mine. May your vessels never shatter
56. (NIH Funding) - Is it possible that there is actually a useful motivation behind this change?
If NIH is mortgaging their grants, then that makes them sensitive to the NIH budget in future years, which is highly uncertain given the current political climate. So, NIH could end up funding the first year of a 4 year project and then needing to cancel it in a future year due to lack of funds. That would then waste 1 year of resources. If they fund the full grant up front, then the money is already effectively spent, thereby reducing uncertainty.
NSF made similar changes about 15 years ago during another period of economic uncertainty, where they deliberately made fewer grants for a couple of years but spent all the money up front in order to reduce their "mortgage" rates. Their intent in doing so was to reduce uncertainty in the budget.
I don't know the details of this NIH change, but it is possibly not nefarious. Of course, this administration has given us little reason to give them the benefit of a doubt either.
Re 52, Amelia (albeit duplicating what I wrote in the hidden thread - 'scuse, I visited there before seeing this post):
It is ironic, given the (valid!) concerns about "superpersuasion" via AI, and the potential for this use to be part of oppression by governments, that the last few weeks have seen a use of AI video generation by the _resistance_ in the UK. It is doubly ironic that the "Amelia" character was originally created by the Starmer government as part of the "Pathways" propaganda game, to disparage unWoke thought. Instead, her widely shared unWoke views lead to her being incorporated in AI-generated video clip after video clip (e.g. https://www.youtube.com/watch?v=xIdSect2cyo&list=RDxIdSect2cyo&start_radio=1 ) and now she is the symbol and avatar of _dissent_ from a government that has arrested 12,000 people a year for speech.
I don't expect this to last - few governments are _quite_ so stupid as Keir Starmer's, but it is a refreshing irony to see this tool used, however briefly, for free dissent.
It's not uncommon for the right, particularly the alt-right, to adopt and reverse media intended to attack them. One of the big weaknesses of the right is that it doesn't really have a pipeline to produce friendly fiction, because the creative professions in the West are disproportionately SJWs - they're not 100%, but they're enough to keep the rest in line via the threat of blacklisting.
The alt-right's solution is to take SJ-promoting media and simply identify with and promote the villains, particularly when that media fucks up its messaging such that normal people find the villains sympathetic.
"Keep Your Rifle By Your Side" is another prominent example.
Many Thanks! It is amazing how successful it has been in this case - and how much it (currently!) differs from the typical AI-as-a-tool-of-tyrants concern.
It seems sort of predictable to me. Democratizing the means of cultural production seems like it would always lead to more dissenting voices being heard.
Many Thanks! But that the introduction of AI video generation wound up democratizing cultural production rather than being differentially available to better funded and/or more technically sophisticated groups, _centralizing_ cultural production, was not widely expected.
Why wouldn't it be expected? The first time I saw Sora I thought "this is the end for video production studios". Making a video like that before LLMs would've cost, I dunno, tens of thousands of dollars? Now it's basically free. How could that *not* lead to democratization?
Many Thanks! That the price drop was as large as it wound up being was not the only possibility before it actually happened. Consider what would have happened if the technology had followed an alternate path where the quality went up (say to routinely photorealistic quality) but the price had dropped only slightly or not at all.
There are additional hazards that might lead to centralization of control that haven't happened yet but might still happen: Consider that there are relatively few frontier AI labs, so they are potential choke points for censorship. As is, there are some restrictions on content that AFAIK, against e.g. CSAM, that are enforced by that set of labs. If, e.g. the UK government decided to ban 'politically incorrect' video generation, they just have to coerce a handful of labs to enforce it. And detecting which prompts fall into that category would use the _same_ language understanding capabilities that make it possible to generate video from a text prompt in the first place.
I'm thankful that democratization happened, and that dissenters have a voice, in fact aided by the technology. I'm not sanguine that this was the only path that could have been followed, nor that it will persist over the long term.
Re: link 51 and GetBrighter, speaking as someone who's potentially interested in brighter illumination sources, the price is a sticking point for me; I'd rather pay less than buy something whose materials are selected for an "ultra-premium feel," and it strikes me as highly relevant that I could get the same amount of passive illumination for cheaper by buying a power strip and a set of conventional sunlamps to plug into it.
I already bought a sunlamp a while ago, have been thinking I'd be interesting in upgrading, but checking out GetBrighter just convinced me that the logical upgrade option would be "more sunlamps."
You can see here[0] for how it compares to alternatives and even DIY'ing, we're pretty cost competitive on lumens/$, especially when taking CRI into account.
It doesn't look like this is comparing to a very wide selection of alternatives- you say that you're "pretty cost competitive on lumens/$, especially when taking CRI into account," but all the alternatives you actually compare on a lumens/$ basis are ones specifically marketed as having comparably high CRI, and while all of them are significantly cheaper than GetBrighter, you can get significantly cheaper than that if you're not looking for options that specifically aim to maximize CRI.
For me, smart home integration has zero value add (if anything, I'd prefer to avoid it,) and I'm not especially fussed over the difference between 85 and 95+ CRI, and with those priorities, I can get similar levels of illumination at roughly a quarter of the price.
Sorry I phrased it poorly. I meant to say we're not too far off lumens/$ vs DIY'ing given we're an actual consumer product (& the features like adjustable cct, dimming, smart home + manual controls). But depends on your threshold for "too far off"...
That's fair. I may simply not be part of the target market for this sort of product, but for what it's worth, as someone who *is* quite interested in high-intensity lighting options, I'm sometimes willing to pay more for better form and convenience, and the higher CRI has at least some positive value to me, but premium materials matter very little beyond the point where I'm confident it won't be accidentally broken, and the smart home connectivity doesn't offer any positive value. I'd be willing to pay up to about half the price of the GetBrighter lamp for something that offers illumination of several sunlamps, shares their adjustability options, but has the added convenience of being a single point source which is convenient to move around the room and doesn't require a power strip or multiple outlets. I have no idea whether exploring a cheaper option without smart home functionality or anything like that is something you'd want to explore, but I suspect there's at least some available market segment to be captured that way.
> Aren’t many murders in very fast attacks where the murderer doesn’t get to choose how many shots/stabs to land?) and would welcome more research on this topic.
I would assume, until I had good evidence to the contrary, that the medical care effect is just not very large, at least not over the timescales referenced here. Have we really gotten significantly better at treating gunshots in the past few decades? (Also, this seems like maybe another way to test the theory--look at murder rate by method, and see if the more treatable ones have seen a bigger decline).
I don’t have any stats handy to prove this, but a pretty large number of murders are dumb impulsive crimes, people getting into a fight and one grabbing a knife or somebody getting kicked out of a party and coming back with a gun. A lot of of those murderers are intoxicated, angry, and not thinking very clearly.
"the National Eye Institute did a study like this to prove that a $50 older drug worked just as well as a $2000 newer drug"
You're killing me, man; the word "prove" is simply not appropriate in this context. It's a bad habit of yours and it makes you look statistically naive. One of the first things I teach students in stats 101 is to not think in terms of proof but think in terms of strength of evidence. The distinction is not meaningless, and I am not an old man yelling at clouds. For example, in case you're not willing to take the words of some rando commenter:
Statistics is fundamentally about quantifying uncertainty, whereas proof exists in a domain with zero uncertainty (e.g. theorems). Again, this is *not* a trivial distinction and in fact is a rather large problem in science communication more generally.
In fact, the meaning you would ascribe to it is not even the most common one: to prove something in a court of law you can rely on evidence and testimony, for example, which I assume your definition would exclude.
I get being a stickler for specific definitions, but the word "proof/prove" is not exclusively (or even mostly) a technical term
Who cares is that's how the word is "commonly" understood? Here it's being used in a specific context, namely a scientific context, in which my assertion is correct. No need for mental gymnastics to defend improper use of a word.
I'm sorry to have to push back on this, but I don't understand where you get your definition.
On the Oxford Handbook of Medical Statistics, for instance, there is no technical definition for the word "prove" or "proof". There are, however, usages of the word you would describe as incorrect. E.g 2.4, in the subsection "When consent may be withheld".:
“One solution in situations like these is for the researcher to decide in advance to offer the intervention to all control group subjects after the trial has finished, assuming that the intervention proves to be effective.”
In section 2.4, about placebos, we find the repeated expression “proven intervention”, in the phrase:
“The benefits, risks, burdens and effectiveness of a new intervention must be tested against those of the best current proven intervention”
How are all these things "proven" if studies cannot "prove" anything? Clearly the word is being used in its commonly understood meaning of: show to be true at a statistically significant threshold.
I get being a stickler for definitions, and maybe your approach would be better for communicating the difference, but that's not how it's being used in the technical literature right now, so I don't think you can be annoyed at people using it in this way.
You argument appears to be "some other people are misusing the word too so it's not actually misused," and then you proceed to extrapolate that to "how it's being used in the technical literature right now."
I'm just going to facepalm and walk away for the sake of my blood pressure, and hope that perhaps more thoughtful people will take a second to consider their own limitations on the subject, and recognize that even very strong empirical evidence still leaves uncertainty and thereby precludes use of the word "proof" or any variant thereof. You know, the way actual statisticians do.
> it would be too much of a coincidence for the (trauma-care-induced) decline in the murder rate to exactly correlate with the (recording-bias-induced) decline in the assault rate... Aaron argues that would-be murderers have adjusted by trying harder to kill their victims
"Wisdom of crowds" hypothesis amongst murderers...
#35: Women are typically said to have high rates of mental illness than men, especially teenage girls versus teenage boys. How much of this is the same reporting bias?
On #35, the finding that liberals report their moods to be equal to conservatives doesn't hold up in other studies. I wrote about this here, where I also presented the results of an experiment investigating why there is an ideological mental health/mood gap in the first place. https://www.richardhanania.com/p/does-therapy-culture-explain-the
As I understand it, people consider Armenians white in most of the US *except* California, which has the largest Armenian population in the country. I doubt there are enough Georgians or Azeri in the US to get much attention in terms of racial classification.
When I was growing up in Connecticut there were three groups in town - Lebanese Christians, (mostly secular) Albanian Muslims, and Portuguese - who were “exotic whites,” meaning white in a racial sense but culturally distinct. In contrast, no one ever considered even the most light-skinned Puerto Ricans white.
> A conventional boat pushes through water, and the resistance it encounters increases exponentially with speed. (Specifically: drag scales roughly with the cube of velocity, meaning that to double your speed you need eight times the power to overcome resistance.)
It's unusual to see a claim of exponential growth followed by an immediate proof that the growth is not exponential.
There's an exponent; there isn't an exponential, which refers to a different type of object.
"Exponential growth" could be interpreted as referring to "growth in the exponent", with "exponential" as an adjective, or as "the rate of growth that characterizes exponentials", with "exponential" as a noun. (Both are defensible under the normal rules of English grammar, and the ambiguity can't really be resolved since those two definitions have identical meanings.)
Your explanation relies on the speaker not knowing any meaning at all for "exponential". This is often the case, but I don't see that your explanation can explain much. You wouldn't see someone confusing the noun "maniac" with the noun "mania", even though the derivational relationship is basically identical to the one between the noun "exponential" and the noun "exponent".
The explanation I'm thinking of is that people are interpreting "exponential" as meaning "having an exponent". They think that linear functions don't have an exponent (because we don't write down the 1), but think of both 2^x and x^2 as "exponential" functions because they have exponents.
After I stopped moaning in pain, I prompted Google search (and presumably Gemini, under the covers) with
>in the US, with normal schooling, at what IQ do half the population successfully learn what an exponential function is?
and it/they responded with:
>Based on IQ distribution and typical educational outcomes in the United States, half the population (roughly the 50th percentile) successfully learns and understands what an exponential function is (typically taught in Algebra II) at an IQ of approximately 100
( I _suspect_ that a higher IQ is actually required... - but maybe I'm cynical. )
In regards to the Helen of Troy one, I am not sure that the Trojan War era Ancient Greeks actually understood how skin color and other "racial" characteristics were inherited. In stories of the Trojan War the Trojan ally, King Memnon of Ethiopia, is depicted as black in Ancient Greek artwork. However, he is also attested to be the son of Tithonus, a Trojan prince, and Eos, the goddess of the Dawn. Neither of these characters are depicted as black in ancient artwork.
The myths do say that Eos took Tithonus to live at "the ends of the Earth," which would imply somewhere in Africa since Memnon grew up to be an Ethiopian king. That implies that the Ancient Greeks at the time of Homer thought that anyone born in Africa would become black, even if they had white parents! So maybe Leda and Tyndareus were on a diplomatic mission to Africa when she had Helen (or rather, laid Helen's egg).
Fun fact: The Prose Edda identifies Memnon as the father of Thor. But also that Thor had 'hair fairer than gold' so presumably Snorri Sturluson wasn't completely familiar with the details of heredity either.
The darkness of Africans was mythologically explained by the sun god's son Pheaton's demand to control the sun for one day. He lost control and scorched part of the earth, creating the Sahara and black people. Perhaps Memnon was alive and in Ethiopia when this happened, and the Greeks believed in Lamarckism, so that his descendents would still inherit his dark skin.
Or they didn't think it through and just thought "every ethiopian king that I've hever heard of is black, if you tell me memnon is an Ethiopian king i'm going to depict him as such".
23. With respect to Lawyers and AI--my evaluation is that the use of AI in my practice (which is primarily regulatory and related to strategic counseling rather than litigation) is that AI is extremely useful, and the utility is not limited to ChatGPT 5.2 Pro or its ilk. A lot of firm work involves very quick turnaround and (for lack of a better word) dumbing down complex regulatory requirements into intuitive, gut level lists of "what you, the client, need to do."
After years of doing this, I usually know off the top of my head what the client needs to do with respect to the regulations that I work in, but I don't always have the exact citation and language off the top of my head. AI is excellent at fetching this from ecfr.gov, providing exact excerpts, and making this digestible for a client that is overwhelmed by regulatory complexity.
While the legal "know-how" that comes from intuition and general market and regulatory evaluation is, as the OP says, the real value that a lawyer provides, there is blackletter that is very much amenable to digestion via LLMs.
I am not a tax lawyer (though I work with them extensively), but I would imagine tax law would also be like this. Any realm of law that has to do with blackletter regulation and which can, therefore, probably be formalized will be the place where AI shines in the legal profession.
I once explained to Claude how I make good coffee in a basic Mr. Coffee. It was surprised it worked. It especially thought that using a microwave to heat the water to bloom the grounds was a bad idea.
Re: 28 (the LLM getting rewarded for using the calculator)
I hear that in some of its training runs, evolution rewarded people for looking at images of fertile young people of the opposite sex, and now lots of humans spend time looking at (or imagining) pictures of fertile young people of the opposite sex even while working on completely unrelated tasks!
38, I read the linked blurb and this seems really confused. As “rights” talk, affirming or denying, tends to be IMO.
A “right”, practically speaking, is just what people and coalitions could make the wider culture they operate in accept as a “legitimate” want. “Power relations” etc.
E.g., people who can acquire some stuff through work and trade are generally more competent (and have been, e.g. in America, been very numerous), so getting to keep what you acquire that way is a “right” to property. Because many competent people would obviously just use violence to come to the same result otherwise, only much less pleasantly.
Inventing a new “right”, always a popular pastime, is just a claim to legitimacy of a want; if people (coalitions) can credibly make it stick, others will agree, otherwise, it’ll be met with something on the laughter…violence spectrum by the rest of society. Mostly, it stays with the middle ground of indirect, regulated, and institutional violence (voting).
Most people really, really want the culture they are used to to pervade their lives. Being a stranger in a strange land at home is disorienting and intensely painful to them. Outright majorities prefer marginal familiarity to marginal disposable wealth (if not, the real estate market would look vastly different).
Trying to keep moving policy in directions that very many, largely competent people deeply disagree with by calling their innermost wants illegitimate (“not rights”) is not likely to make people reconsider what they want, it’s just fighting words. So far the debate has escalated to “Trump”. It puzzles me that some people want to escalate it further; people will not want differently if scolded more.
It seems likely that improved trauma care could reduce the amount of (successful) murders by SOME amount.
It does NOT seem remotely likely that a 2x improvement in trauma care would consistently reduce murders _by a factor of 2_ (unless you quantify trauma care in a way that makes this tautological). The relationship between them seems obviously far less direct than that.
I wish people would put more thought into the actual quantitative relationship between the lever they are pulling and the outcome they are measuring.
Re 36 rooftops houses, we have some of these in San Francisco at Embarcadero Center. I even know someone who lived in one. They said it was reasonably nice.
However, WTF is up with the roads, cars, and presumably garages. Who is driving their car up there, how (10 minutes on ramps?), and why (the building doesn't have an express elevator for the luxury homes at the top?).
My take on the lab leak change is that (1) it was a believable story that never had any solid evidence and (2) the X-is-destroying-America mob have moved on from China to maybe AI, Trump, America, or something else. The wind only blows in one direction for so long.
Thank you. I’ll adopt the same translation and it will save me frequent momentary frustration. But put me down in favor of folks just using cite signals the normal way.
#38: The word 'populism' has always confused me. Appealing to the mass of people with issues that they care about, in the hope of getting votes from them - this is just democracy working.
"But the mass of people doesn't even know what they want, they will vote for something that sounds good but turns out to be bad!"
Sure, and so says every out-of-touch, insulated and arrogant beaurocrat ever. History abounds with examples of people saying that the common people don't really want {x}, e.g. look at the speech "Slavery a Positive Good" (https://teachingamericanhistory.org/document/slavery-a-positive-good/) which argues that slavery is Good Actually and that the Mass of People who call for Abolition are Ignorant and Thoughtless and this will lead to the End of the Union.
Is the desire for regime change in Iran a 'populist' movement? If not, why not?
If so, is this good or bad?
Does this mean populism can both good or bad depending on what policies are being pushed (good populism, bad populism)?
If so, why is the word 'populism' overwhelmingly used negatively?
It's populism if it sounds good at first blush when you don't think about the consequences or have little knowledge on the topic. As most people don't spend much (any) time evaluating policies (very much separate from yelling about politics), the name fits.
The prototypical example is lowering taxes. Those who support it often didn't go over the financial plan of the country, they just see that their personal expenses will decrease on that one item and so they support it.
Obviously, when someone claims something is populism, they can be right or wrong, it's not a magic word. So someone claiming abolishing slavery is populism was simply wrong.
And you don't know whether you are right or wrong until much later.
So lowering taxes is populism, and bad.
However, if it turns out that lower tax rates leads to more economic activity and more wealth for all, then lowering taxes is retroactively good, and therefore it can't have been populism.
(The Laffer curve - some tax rates are just too high.)
It seems that calling a policy 'populist' is a slightly fancier way of saying 'I think it's bad'. You don't actually know the outcomes until they play out.
Sure you can predict and forecast and go off prior examples - but there is uncertainty in all of this, and that uncertainty and humility is reflected in the statement 'I think it's bad' but not in the statement 'that's populist so it's bad'.
It's more like "It's bad, but it seems good on low-informatiom.".
I don't see a problem with having a succint word for this and how much humility a speaker has seems nigh orthogonal to me and mostly depends on context. For example, one can also say "I think this policy is populist".
Even granting that, using populist for this is bad, because it causes confusion and makes the speaker look like an arrogant elitist who believes that only the highly credentialed can have meaningful opinions. When many "elite" opinions fit the definition way better. Including woke, DEI, and many other things.
Nah, I could have told you abolishing slavery was a good idea even before you Americans did it. Even if there had been severe economic implications, the inherent value of freeing people would have still outweighted it.
Prior to the Industrial Revolution, free societies tended to be conquered by extractive premodern states which enslaved or at least villeinized the majority of their population, and were able to raise professional armies on the base of that unfree labor.
The fact that liberal societies are viable today rests on our ability to outsource hard work to machines.
"It's populism if it sounds good at first blush when you don't think about the consequences or have little knowledge on the topic."
Oh? So if you do think about the consequences and decide that you're fine with them even if it makes your opponents extremely angry, it's no longer populism?
Take Brexit for an example: most people who voted for leaving the EU said they were fine with taking a moderate mid-to-long term economical hit if it meant clawing back control over their own laws and borders and ultimately throwing out all the brown immigrants. It didn't seem to stop any remain voters from calling Brexit populism, but you would argue they were wrong, then? Those people were simply seething losers?
I doubt the median voter actually conceptualized what moderate mid-to-long term economical hit and lower immegration entailed, but if they did then yes, it would have been wrong to call it populism.
I think part of the distinctiveness of popularism is that it is more explicitly anti-elite than just pro-popular stuff: it's not a focus on popular stuff in general, it's a focus on specifically the popular stuff that the elite don't like--sometimes maybe even conflating "not liked by the elite" with "popular in general" in a way that isn't justified.
Whether "populism" is reliably used to single out movements with this sort of inclination I'm not sure, but IMO that's a meaning of populism that I think distinguishes it from just democracy in general, and also one that makes it clear what the downsides could be: an inmate mistrust of elite institutions might mean a bad information ecosystem; mistrust of elites as a class means it's hard to accomplish things through regular political channels, since the will usually require collaboration or compromise with elites--meaning that there's a tendency to either diffuse but unproductive rage, or radical destabilizing change; etc.
Of course, a lot depends on what "elites" a movement is suspicious of: I'm more sympathetic to middle class lawyers and doctors being mad about kings and noblemen than I am to the "lumpenproletariat" of gamblers and mobsters being mad about judges and scientists, but of course YMMV.
"If so, why is the word 'populism' overwhelmingly used negatively?"
Because the professional middle class controls the means of communication and their views are at odds with those of the mass of the people on almost every point. Even their Fabian socialism is something they want to force onto the common man top-down for his own good. Chesterton wrote very lucidly about this already 120 years ago and nothing has changed. Orwell is also good on the problem of middle-class socialism, especially in Road to Wigan Pier.
It seems to me that you (and many people) are conflating a bit the meaning of "populism" with that of "demagoguery". The later means using a rhetoric that has a vast appeal using the prejudices of the mass instead of their rationality, and has a pejorative meaning per se.
Populism means using a rhetoric about (sometimes pretending) favoring the well-being of the "common people" to that of a (sometimes chimeric) elite. To me it is not derogatory per se, as there are indeed many society where there is an elite who favors its own interest going against that of the mass, for example indeed Iran. But in other context, this dichotomy between the mass and the elite is blurry, imaginary or cherry-picked (is the elite really the intellectual left rather than the billionaires?), or/and is applied to societies where such opposition wouldn't provide any benefit to the "common people".
Of course, demagoguery is often used in populistic rhetoric.
Microsoft was founded in Albuquerque (see, e.g., "History of Microsoft" on Wikipedia). Gates and Allen dropped out of Harvard and moved to Albuquerque because the original plan was selling a BASIC interpreter to MITS, which was based in Albuquerque. The company was founded in 1976 and had 13 employees when it moved to Washington in 1979.
I judge the rest to be accurately West Coast, subject to the proviso that the Aramco thing is a bit murky among other reasons because Standard Oil of California was one of the successors of Standard Oil, which was founded in Ohio. Obviously Standard Oil's west coast acquisitions were originally founded on the west coast, but still...
The recent furin cleavage site (FCS) finding in Brazil is of no importance. We already knew many betacoronaviruses have FCS. What we've never seen is a sarbecovirus FCS, and this evidence has only become stronger as more sarbecoviruses are discovered and still no FCS.
And of course, this rare FCS happens to appear as an insertion (very rare), and with the CGGCGG sequence that is nowhere else in the SARS2 genome.
"39: When complaining about modernity’s real and obvious flaws, it’s important not to forget how much lots of traditional societies sucked: “An Egyptian Muslim woman who lived under female seclusion since her marriage, 40 years ago, asks a female Christian missionary to describe flowers to her.”"
That's not a flaw of traditional societies. That's a flaw of fundamentalist Islam. Even under the Roman Empire, Christian societies never practiced female seclusion. In the 21st century, many Islamic societies still do.
> I fact checked myself and Saramago's novel is not actually set in the middle ages, nor even the early modern period but, as it's the 18th century, it's the Age of Enlightenment.
Ah, although I suppose in terms of what you were talking about, nunneries, not much had changed.
> Granted Saramago's book is fiction, not History.
I'm not sure whether Saramago would consider nuns living in a nunnery as an example of societies practicing female seclusion. Would he have considered male monks as an example of male seclusion? If both monasteries and nunneries are considered as examples of seclusion in a society, can we say that society is characterized by female seclusion?
Also, more to the point, I believe that in Portugal - like in most medieval Christian nations - daughters were given a dowry at marriage. After receiving a dowry a daughter would have had no claim to an inheritance.
Not only are there flowers in convents, there are famously large quantities of them. Also, the type of women who in our day are vehemently feminist entered orders voluntarily in the middle ages, so it's not necessarily straightforward to distinguish between "seclusion" and "separatism".
There's something very suspect about the World Happiness Report that I never got around to fully investigate. When you actually ask people directly whether they're happy, China wins by a margin [1]. Most notably, they're 83% positive on "country's social and political situation", compared to a world average of 40%. Interestingly, they stopped surveying China since.
This is a result that is fairly risky for a western institute to publish, and I always suspected that's why WHR engineer these weird questions that somehow bring Scandinavia on top. But I'm willing to be convinced otherwise.
Note that the common western view that "China threatens people to answer positively" is a silly myth not supported by evidence. Chinese have no problem telling pollsters that corrupt officials are the country's main problem - a much more sensitive issue than saying "I'm not happy". [2]
Do you have a source for China not being included in surveys? One of the authors is from the International Business School Suzhou, Xi'an Jiaotong-Liverpool University, which has a location in China. While I've only just skimmed the info, I see China mentioned in several chapters from the 2025 report. And looking at the data it seems like it's included.
Is Ipsos affiliated with the WHR mentioned in the OP links?
Skimming 2020-2025 WHR ranking data, I don't see China coming in the top 10 in any of the years. So it seems strange to me that they would juice their metrics to prevent something that's never happened in their previous reports.
As for Ipsos, I noticed that while China wasn't included in 2024 or 2025, though in 2024 at the bottom of some slides they state "The samples in Brazil, Chile, China, Colombia, Indonesia, Ireland, Malaysia, Mexico, Peru, Singapore, South Africa, Thailand, and Turkey are more urban, more educated, and/or more affluent than the general population."
That makes it seem like they had the China data, but didn't include it, or maybe just copy-pasted from a previous year's report.
Ipsos is a separate survey that asks the direct question. They're not related to WHR, who ask different questions that put China lower.
I didn't notice the comment you provided. I agree it's likely a copy-paste error in 2024.
Like I said, this is just some suspicion I have. I never really dug into it. In any case, Ipsos stopping surveying China is less important, since just the 2023 survey is enough to raise the suspicion that WHR (the recognized and respected happiness index everyone quotes) are hacking their results.
> I thought the commercials were in bad taste, misrepresenting what OpenAI’s ads would be like and turning the completely normal decision for a tech company to have an ad-supported free version of their product into some kind of horrible betrayal.
En contraire I thought it was a refreshingly honest statement (for a company) about the, pardon my lingo, surveillance capitalism that has been the primary business model for big tech of the last decades. Google betrays us by trapping the less tech-savvy (aka average) user into clicking on ad results. Meta betrays us by luring us with connections to our community while hooking us into content loops that leave us passive, overly-activated and eager to buy shit we don't need to give us relief.
I might be overstating and overdramatizing the case a bit, but I think the sentiment is growing and there is a lot of truth to be found in there. Add to that the alignment worries, with the misaligned over-optimizing for engagement machine that capitalism already is, even without viciously fast AI self-improvement iterations, and it looks like a very reasonable ad.
> I thought Sam Altman’s response was fair (although his countercriticism of Anthropic also missed the mark). People in his replies tried to enforce a norm of “if you write a long explanation defending yourself against someone else’s funny lies, that means you care and you lose”, but that’s a stupid norm and people should stop shoring it up
I only skimmed the comments on https://xcancel.com/sama/status/2019139174339928189, but they seemed more the usual level of X quippyness/brain-deadery to a tweet that read butthurt. At least from the view I got on it, I couldn't see any norm-enforcing (except for, don't be a bad loser)
Framing someone correcting implied falsehoods about their company's advertisement operations as being a bad loser is exactly the norm-enforcement Scott is talking about.
I gave arguments for why they are not falsehoods, his Alt’s post didn’t read as correcting anything to me, hence the only norm I’m enforcing is not being a bitter loser
"DC has a rideshare app called Empower that charges 20-40% less than Uber. (Drivers like it too because they keep 100% of the fare)...DC is trying to shut it down because of liability insurance. DC law requires $1 million per ride. The $1 million requirement isn’t sized to typical accidents. When $100,000 is the limit available for an insurance claim, 96% of personal auto claims settle below $100,000...Empower can offer $7 rides partly because it circumvents the mandate. DC is shutting it down for exactly that reason.”
8 - i am not a christian and have very few friends who are but the smug self-righteousness that non-christians point at christians bugs me. People who bend over backwards to avoid racial discrimination exhibit joyful religious discrimination. Who cares if the characters in a christian video can't be christians because they are not human. The vast majority of christians are just trying to be good people i believe - just leave them alone.
I'd say that my experience is the opposite--most of the "anti-religious" sentiment I've seen was, when you looked closer, just hated of Christianity. Including most capital-A Atheist discourse. They had no problem with whatever half-understood pseudo eastern woo was currently fashionable. In fact, if you repackaged boring Christian thought into an eastern (or at least non obviously Christian) packaging, they'd swallow it whole.
> 8 - i am not a christian and have very few friends who are but the smug self-righteousness that non-christians point at christians bugs me.
Maybe you're just in the wrong circles, or noticing the wrong people? Within our own monkeysphere, you have the Psmiths, who are both Christian and write impossibly interesting, erudite, and articulate book reviews (as a couple, no less!), and which have been in Scott's blogroll and whose posts are commented on in open threads here.
When I worked in finance, one of my good friends was a Christian. Most of us were Phd's, most of us were data science or modeling or adjacent, and he was one of the smartest and most thoughtful among us, as well as a genuine role model in all he did. And as far as I could tell, everyone else respected him just as much as I did.
The book Bonds That Make Us Free, by C Terry Warner, is recommended by CFAR, and is a fairly religious book by an openly religious author.
I feel like I see numerous examples of people engaging thoughtfully and respectfully with Christians in our set.
Thanks for this. I did not mean to imply that everyone here has that smug attitude when it comes to christians. I did mean to imply that item 8, the one pointing out that a christian video series, VeggiTales, is stupid because the characters are not in fact christians, seems to me to exhibit the small-minded tribal prejudice against christians that i see in our society. I don't like it.
Did we read the same blog post? I parsed it as a respectful, nerd-enthusiastic account of how the show's Christian creator had in fact been remarkably consistent in his application of his own dogma, contrary to the claims of his (less insightful, also Christian) detractors who were calling the show blasphemy.
I'm afraid it took me too long to read the link and now most people will have read it, but I would really like to push back on this. The linked article is Not Very Good, and the article it's responding to is at least a bit more interesting.
I'm saying it's "Not Very Good" because it doesn't address most points raised in the original, and it has an annoyingly dismissive tone; like when it accuses the original article of being "a little rhetorical" and that the audience "responded to trigger words like 'elite misinformation'".
The first one is annoying in its own right, since the tone of blogs is often halfway between the informative and the personal; I'm not surprised to see references to personal opinions or perceptions of a matter, especially when they serve as hooks to discuss the actual analysis.
The second one is annoyingly taken out of context, and reading where it appears in the original gives a much different impression of the piece. The quoted section appears in the concluding paragraph:
“Over the last years, media outlets like the New York Times, universities like Oxford, and international institutions like the UN have devoted themselves to the fight against so-called “misinformation.” It is certainly true that our political discourse is awash with dangerous distortions and outright lies. But any institution which wishes to address that problem must start by looking into the mirror—and cease spreading “elite misinformation” like the World Happiness Report.”
The author is drawing a parallel between the scaremongering about "misinformation" done by traditional information institutions and their own irresponsible spreading of half-truths, which has been proven (according to the author at least) in the rest of the piece.
It's needlessly inflammatory to simply quote the part about "elite misinformation", which elicits the image of a conspiratorial mind, when the author is actually accusing traditional outlets of writing clickbait, which is a much more reasonable accusation.
Yascha doesn’t accuse traditional outlets of writing “clickbait;” he accuses them of spreading “misinformation,” which he seems to equate with “dangerous distortions and outright lies.” This is known as “poisoning the well;” by declaring the World Happiness report to be misinformation and a sham, he is preemptively declaring that anyone who claims to disagree with him is lying. Huw has one paragraph where he (charitably, in my view) says that Yascha’s post is “a little rhetorical,” but he otherwise attempts to address the substance of Yascha’s and ignores the rhetoric.
I suspect that Scott linked to the article in part because it was a high effort response, by which I mean that Huw identified Yascha’s major arguments in support of the thesis that “the world happiness report is a sham” and responded to all of them. You write that Huw, “doesn't address most points raised in the original,” but rather than list the points yuou think that Huw failed to address, you devote all of your seven paragraphs to complaining about the tone of a single paragraph by Huw. So we will have to agree to disagree on this.
That's fair, I didn't get to the substance because I was put off by the tone of the reply. But I would still strongly argue that the posted reply doesn't do justice to the original piece at all.
For instance: the original points out that different researchers publishing happiness scores can arrive at very different scores for the countries involved, despite ostensibly measuring happiness. If you had two IQ tests that gave out very different results, you would rightly be skeptical of the efficacy of the measurement.
I guess in part I am mixing in my own opinions about the topic, because Yascha's article doesn't spell this out in so many words, but it's clear from his bringing up a different happiness score that gives different results as a counterpoint to the validity of the World Happiness.
Also, this is in the reply article:
It is not clear to me that finding a bias in the ladder contradicts its purpose. Is it not reasonable that having money and being successful might make people satisfied with their lives? (Again, the individual-level results in Tables 11–13 find that life satisfaction correlates with income, and also age, gender, freedom, marital status, social support, health, institutional trust, etc.).
This is clearly circular reasoning: bias in the measurement is not a big deal because... the measurement is found to correlate with the items it is biased in favor of?? Obviously if you skew the questions to give more weight to economical factors it will correlate with those; same for the other biases.
I reread the original article and I concede that the tone can be annoying, especially if you disagree with the premises, but the answer isn't really much better, complaining about the tone of the original while making logical mistakes and just shrugging at the valid concerns.
(concerns much better expressed in that one _sam[ ]zdat_ post)
I went back again to the original article, but I'm starting to suspect we are reading two different pages. The one I read has the key takeaway in the last two paragraphs that read:
More broadly, supposedly serious news outlets still have a long way to go in subjecting publicity exercises like the World Happiness Report to appropriate journalistic scrutiny. It is easy to see why ***editors are tempted to assign some beat reporter without expertise in the social sciences to write up a fun little story about how much happier those enlightened Scandinavians are compared to benighted Americans.*** But if the media wants to live up to its self-appointed role as a gatekeeper of reliable information, it can’t continue to be complicit in the spread of such ***shoddy clickbait.***
Over the last years, media outlets like the New York Times, universities like Oxford, and international institutions like the UN have devoted themselves to the fight against so-called “misinformation.” It is certainly true that our political discourse is awash with dangerous distortions and outright lies. But any institution which wishes to address that problem must start by looking into the mirror—and cease spreading “elite misinformation” like the World Happiness Report.
(apologies for the lack of formatting, I don't know how to highlight stuff)
Again, it is clear from the phrasing that the accusation is that journalists are tempted to write according to simple narratives rather than research a topic accurately; this lack of research shows up as shoddy clickbait; again, the point made in the piece is that the outlets that pick up the news are latching onto the idea that the US is falling behind on "happiness", and should learn from northern Europe. It's not hard to see the motivated reasoning behind this choice.
We don't really have to agree to disagree on the specifics though: the response post addresses none of this, it just defends the measurement as flawed but useful, which is tangential at best to the original post. It is high-ish effort, but it doesn't identify the major arguments nor does it address all of them (at least not satisfactorily; see the logical contradiction in hand-waving away the reported bias in Cantril ladder measurements)
My mistake; I reread the last paragraph before posting but didn’t check the preceding which talks about clickbait.
Huw is writing about the thesis: “The World Happiness Report Is a Sham.” That’s the title of Yascha’s piece, so Huw is not nitpicking by focusing on it. He ignores the media criticism except to say that it doesn’t support that theseis: “The WHR are extremely clear throughout their report about their methodology, and present explanations of it in a long preamble before presenting their annual table; it is simply not their fault if media outlets looking for a story obscure this.” So, yes, there is a bunch of stuff that Huw does not address.
With regard to Yasha’s contention that, “the Cantril Ladder...does not even do a good job of measuring respondents’ satisfaction with their own lives,” his support for this is research that concludes that, “the Cantril Ladder’s structure appears to influence participants to attend to a more power and wealth oriented view of well-being.” Two problems. First, “appears to influence” doesn’t support a definitive assertion; at best it would support “appears not to do a good job.”
Second, the implicit claim being made here is that, “a more power and wealth oriented view of well-being” is a poor measure of well-being. Huw disagrees. His position is that, “self-reported wellbeing measures are accurate by definition,” which he follows with a bunch of qualifications. I don’t find that completely convincing because it’s defining the problem away. Huw allows that results thrown off by an anchoring effect of previous questions. It’s not clear why that is different, in principle, from the wording of the question itself throwing off the results.
However, Yasha made the implicit claim, so the burden of proof is on him. He provides no argument, convincing or otherwise, to support it.
I'll update my belief that the original is less-than-fine, but I still think the answer is a bit lacking, although you raised some good points about the prominence of the sham-conspiratorial angle.
If you're interested in reading more about the topic, though, I have to recomment the sam[]zdat article I mentioned before. I think you might find the style extremely grating, so it's possible you'll deeply dislike it, but I find the author's position to be strongly argued and well supported. Cheers!
#49: I wonder if this story is quite right. The initial SF Chronicle piece that broke the Ellison yacht story at least implied that the name came from the previous owner, which Ellison kept until realizing the name issue: "1999-11-05 04:00:00 PDT NORTH BAY -- Sailors who hang around the Sausalito harbor were admiring Larry Ellison's newest yacht, the $25 million "Izanami" docked at Schoonmaker Marina. Noticing that the 192-foot boat was made in Germany, they were somewhat shocked when they figured out what "Izanami" is backwards.
But we learned that the previous owner is Japanese and that "Izanami" is a goddess in Japanese mythology. Anyway, Larry, if you want a new name, we suggest "El Caro," the expensive one. It works forwards and backwards."
The brief mention in "Softwar" does not clear this up
I wonder if historical boat naming records are available to confirm or disconfirm this
"I’ll take this opportunity to pitch my startup idea - a dating site where, instead of checking boxes to see if you match, you give a willingness-to-date between 0 and 9, and match if your combined WTD is 10 or greater (so it could be both people rating the other 5, or you rating them 9 and them rating you 1, and so on)."
I like this idea but worry it would fall victim to "rateflation" in the same way, e.g, customer service ratings do, where 5 stars is the default. I understand the incentives here are different since votes are (essentially) private (although if you rate someone a 1 and you match, you know they rated you a 9), but I still think most people are just uncomfortable with eating people they like low, or the general idea of putting how much they like someone into a number. Not to mention if you just rate everyone a 10 you get more matches, which incentives desperate people in the same way other systems do. Also generally people probably want to match with people who like them about as much as they like them, not huge asymmetries. Probably a workable concept with some iteration, though!
Banned for contentless nasty tu quoque comment.
Used to be that when you banned someone for a comment it was possible for the rest of us to read the comment. Now it isn't. Did you decide to change that? I always found it interesting to read them.
I think that's still the case by default, e.g. (https://www.astralcodexten.com/p/links-for-february-2026/comment/210358674)
IIRC the moderator has the option to choose whether the comment is deleted or if it is merely hidden. Looks like scott decided the former for this one, but not for the one breb links below
Thank you for banning some people, seeing the deleted comment was the reason I checked out the comment section again after mostly ignoring it for a few weeks.
I also asked Claude Code to create me comment fact checking extension for Chrome since the leftover un-sourced lying comments were still annoying.
> So, reportedly the segment was nixed for containing no new information whatsoever and therefore being pushed out in favour of news[...]
Well, a major factor was also apparently that the administration refused to comment.
> since none of this is news and everyone knows about this prison
60 Minutes is apparently the highest-rated broadcast (not cable) TV news program and has been for a long time (although this scandal seems to have hurt them a lot, at least in the short term).
https://www.paramountpressexpress.com/cbs-news-and-stations/shows/60-minutes/releases/?view=111479-60-minutes-closes-out-2024-25-season-on-top-as-americas-1-news-program
What "everybody knows" is going to seem pretty different to different people. Covering something extensively on 60 Minutes is how your grandma learns about it.
> [...] as in "probably" the Trump admin put pressure on 60 minutes not to air the report [...]
Well, maybe not in the sense of somebody calling them up and saying "don't run this" - as noted above refusing to comment originally provided part of the justification:
https://www.nytimes.com/2025/12/21/business/60-minutes-trump-bari-weiss.html
Although Stephen Miller had a pretty aggressive response to the meta-story:
https://thehill.com/homenews/administration/5662062-stephen-miller-cbs-news-60-minutes-producers/
But more generally Bari Weiss was installed at CBS by Trump allies and Trump himself has repeatedly praised her because he obviously understands her to be in some sense more favorable to him.
> [...] deporting violent criminals to this prison [...]
16% of the people deported to CECOT had ANY criminal convictions; 4% of the total were convicted of violent crime.
https://www.cato.org/blog/dhs-doesnt-list-cecot-deportees-its-worst-worst-data
I think this further undercuts your point about what "everybody knows".
Also your position basically suggests that nobody will ever care about the truth again and all news reporting is utterly pointless. At the very least I would suggest the Alex Pretti shooting as a strong counterexample.
Also even if you WERE correct, the Trump administration would have to believe you, but they seem very image-conscious.
>> "Bari Weiss was installed at CBS by Trump allies"
> Isn't this baseless hearsay?
No, the Ellisons control Paramount Skydance and their relationship with President Trump is pretty well-documented, e.g.:
https://www.theguardian.com/us-news/2025/nov/20/warner-bros-discovery-takeover-paramount-skydance-larry-ellison
> The criteria was TdA membership.
Sure, that's what the Trump administration claimed.
But:
1. I'm inherently VERY skeptical that the administration was correct in all of these cases or even tried very hard to establish this
2. It's also suspicious that a bunch of gang members have no criminal convictions
3. Kilmar Abrego Garcia was in this group and was not accused of Tren de Aragua membership (he was accused of MS-13 membership, but on shaky grounds)
I think it wasn't an MS-13 tattoo, and CNN found multiple experts who agreed that he did not have MS-13 tattoos.
https://www.cnn.com/2025/05/01/politics/abrego-garcias-tattoos-explainer
> Would you say everyone at CBS and Paramount is loyal to Trump because of this connection?
Clearly not, given the lively internal debate around this story.
> If not, what's special about Bari?
Several things.
Most of her career is a "I'm a liberal, but actually..." shtick where she ends up taking conservative positions and presenting them as either common sense or Things They Don't Want You To Know.
She just got elevated to this position - one for which she's not particularly qualified - by allies of the President.
The President specifically praised her, and he generally has a very strong tendency to praise people on "his side" and denounce people "not on his side", regardless of any other qualifications:
https://www.cnn.com/2025/11/02/media/trump-cbs-60-minutes-norah-odonnell-ellison-bari-weiss
I generally found her explanation for why she delayed airing this story weak and contrived.
> You think there's a reason he had an MS-13 tattoo despite not being a member? Or you think it wasn't an MS-13 tattoo?
The latter, he has some finger tattoos that people claim spell out "MS 13" as though that's an unambiguous interpretation of "marijuana leaf, smiley face, cross, skull" and those aren't just extraordinary common symbols to get tattooed on you.
https://www.cnn.com/2025/05/01/politics/abrego-garcias-tattoos-explainer
Apparently half of my barbers are in violent gangs (admittedly I do select them by looking for people with a lot of tattoos).
Gang members routinely have "no criminal convictions" (In America, this generally means they were underage, and haven't done "anything too serious" to get charged as an adult).
Okay. I stand by my general point that it's inaccurate to call people violent criminals when they have no criminal convictions. If the government wants me to consider them violent criminals, they can prove it via the generally-accepted mechanism of trying them for a violent crime in court.
So, instead of just "shaking the jails and sending the ne'er-do-wells home" you're saying we need to do the full megillah? Are you willing to pay for $300,000+ worth of court time? Including appeals?
https://legalclarity.org/how-much-does-a-murder-trial-cost-the-state/
Please bear in mind "innocent until proven guilty" is a high standard, and that deportation is generally not held to that standard, ever, because we're generally just "sending them home" (and with a nice sendoff song involving marijuana too! La cucaracha).
Trump's gonna say "It's not my fault these folks are being treated so poorly by another country. I run America, not Venezuela or Belize." (perhaps in a larger sense it is, in that he isn't providing them an outlet to escape).
I've put a link to a longer discussion of this upthread in the original post; you can look at the comments there and see if you agree.
My impression is that new boss Bari Weiss wanted her stamp on it. So it was never "suppressed", it was merely delayed with the intent always being to air it eventually.
Does anyone have coordinates or an address for link #36 about the suburban neighborhood on top of a multi-story building that allegedly exists in Indonesia?
One of them seems to be www.google.com/maps/place/The+Villas+(MOI)/@-6.1571015,106.8922891,688a,35y,39.25t/data=!3m1!1e3!4m6!3m5!1s0x2e69f541055c4a0f:0x1e1fb7a71c1a22af!8m2!3d-6.1509571!4d106.8924369!16s%2Fg%2F1pp2vjg6h?entry=ttu&g_ep=EgoyMDI2MDIwMS4wIKXMDSoKLDEwMDc5MjA3MUgBUAM%3D .
Thanks. That's really cool. I wanted to get more angles, and kind of explore it.
In central Chicago there are now a number of examples of that concept at smaller scale. My wife and I viewed several such "townhomes in the air" in the South Loop area about 12 years ago, and had one of them on our short list, when we'd decided to buy in the neighborhood. (Which we did but it turned out to be a traditional townhouse at ground level.)
not quite the same but i've always love this East Village cottage on the roof of an apartment building: https://www.curbed.com/2024/01/seaside-cottage-east-village-rooftop-for-sale.html
The trendline from Kingfisher Towers to this is clearly headed toward cyberpunk favelas encrusted atop arcologies.
I found this:
https://www.theguardian.com/cities/2019/aug/05/suburb-in-the-sky-how-jakartans-built-an-entire-village-on-top-of-a-mall
Thanks.
Seems to be the one.
> Another list of doublets
Also naive and native (but I think this was a slightly different story — the natural phonetic development vs. learned borrowing from Latin thing happened in French already; English just borrowed both)
I don't think #33 is a good way to build intuition re Russian invasion of Ukraine because
1) It is not true, actual whys have to do with its internal affairs and Putin's regime specifically.
2) It sort of justifies horrendous violation of international norms and the bloodiest European war since 1945. Speaking of which, remember the reasons that Hitler used to justify his aggression, would you recommend them "to build intuition" as well?
I don't think building intuition about why someone would want to do something is equivalent to justifying it. I think a map of German population in the Sudetenland would successfully help explain why Hitler wanted to annex it and be a useful historical contribution, without suggesting that wars of aggression are okay.
But isn't it clear that German population in the Sudetenland was merely a pretence? And when you say that this pretence is "building intuition about why someone would want to do something" you are indirectly white-washing them, because you invite us to think of them as a reasonable person that has reasonable arguments instead of a murderous psychopath that makes these arguments on the fly.
I don't think the pretense-real-reason distinction is as strong as you think. I think you can be an aggressive warmonger, and your choices of which aggressive war to pursue will be based on various ideas about national glory and revanchism and how to get your way, and those will depend on facts of history and demography.
Putin claimed that a meme posted on an American Website that just happened to be based in the Ukraine was part of the justification for invading the Ukraine.
Either that's some really good pretense, or Russian intelligence was really off its game.
That sounds like what I've heard about Russian Intelligence.
Did you hear how America/Western Europe spent 10 years wetting their pants about the USSR invading past the Iron Curtain*, when the USSR couldn't do more than drive a few rusty tanks around (and was doing so, in order to attempt to intimidate a West that they very well knew they couldn't fight?)
*absolutely certain it was coming "any day now"
It seems like America cannot get a good read on Russia. This last war (Ukraine, currently still limping along fighting) America thought Russia was strong militarily, but very weak economically (so they thought sanctions would crush Russia in 2 months. Years later... Russia's held strong).
I agree, the casual chain from perceived loss to revanchism, to fascism, and then to external aggression has some explanatory power, both for Hitler's Germany and for Putin's Russia. Although I think that it is important not to skip the intermediate steps in the explanation and to make distinction between reasons as social forces, dictator's personal reasons, and reasons that they present to us as justification.
Also don't forget that people tend to believe their own propaganda eventually. It is hard to live with cognitive dissonance. So at that some point both Hitler and Putin thought that liberal democracies are basically a bunch of sissies who don't have a real man's backbone.
One should always start with the idea that someone is a reasonable person with reasonable arguments (like, say: "mandatory gym class" to cite one policy proposal from Mein Kampf).
Only when they provably make very stupid decisions, like "Protect the coke factory at all costs!"* (despite it being of next-to-zero strategic importance, as it doesn't make end products, and the factories that do make the end products are destroyed), ought one to consider whether or not they actually do understand that coke doesn't always mean cocaine.
*Not actually a statement made by Doctor Rockzo
Why should we think Sudentenland was a pretense? It’s obvious that Hitler wanted Germans in Europe to be part of a united Germany.
We should, because shortly after Hitler invaded many other countries that barely had any Germans living in them.
He had other goals but that doesn’t mean that it was a “pretense”. Everything we know about Hitler suggests he was strongly motivated by the idea of a united German people. There’s no reason to think that this wasn’t the case.
Hitler was a White Nationalist from Austria. That he believed other things shouldn't make you discount this.
I mean, seriously?
That Hitler intended to invade other countries was an obvious fact that Neville Chamberlain understood (and started rearming).
He wasn't a "white nationalist", every European country was white, he was a GERMAN nationalist. Slavs (such as Poles) are definitely white, but they weren't included in his nation.
Because the Germans needed lebensraum.
Yes, "lebensraum"/expansionism is not a pretence, but a true goal.
He had multiple reasons for invading. For example, Norway got invaded because Hitler (rightly) expected the UK to violate its neutrality and take advantage of its coastline, Hitler jumped the gun and grabbed it first so he could protect ore shipments from Sweden (which he didn't have to invade).
Poland was next, and Poland was also a country with a big German-speaking area.
And then all subsequent invasions starting with Denmark came after WW2 had kicked off in earnest and were intended to outflank or defeat France and Britain.
(Again, acknowledging that these things happened for some actual reason doesn't make them okay.)
Poland did, as for France, they declared war on Germany, not the other way around, Greece and Yugoslavia was Mussolini's stupidity, so the only ones that really count are Denmark, Norway and BeNeLux, now those are truly hard to understand. I think Hitler did not simply want Germans in one Reich, but rather "Aryans" - this is clear from the table talk book, that nationalism was really just public propaganda spread to placate the conservatives, and the real goal was always racism.
The German population in Bohemia and Moravia was quite significant, had lived there for 1000 years, and under Habsburg rule had generally happily considered themselves ethnically "German". After Austria-Hungary collapsed, the Germans in Sudetenland assumed that of course they would be joined up with the other Habsburg Germans in Upper Austria, Tyrol, Carinthia, etc. in a state called "Deutschösterreich" ("German Austria") and were quite shocked and angry when that turned out not to be the case. Hitler grew up in an Upper Austrian town right across the border from German majority towns in Bohemia, and seems to have had Bohemian ancestry. So no, I think this was a cause that was actually very dear to his heart and that incorporating Germans that had been "unjustly" cut off from the homeland by Wilson's arbitrary borders was sincerely one of Hitler's main goals. Incorporating Sudeten Germans into the Reich wasn't a pretence, it was one of the issues that radicalized him in the first place.
Hitler had no issue writing off South Tyrol though.
The Alps are a rather big obstacle.
The German population of Czechoslovakia were very gung-ho about joining Germany.
Almost everybody is a reasonable person in their own version of the story. I really recommend reading Gaetano Mosca here. Basically, feeling like a bad guy is psychologically unbearable, everybody comes up with a story that makes them look like the victim or the hero, even someone like Hitler or Putin. There were genuinely good people who were still faithful Nazis: https://en.wikipedia.org/wiki/John_Rabe I know this is hard to imagine, but these were the facts.
Respectfully, I think you are wrong here and wrong to signal boost that info graphic.
This map and the Sudetenland example build sympathy, not strategic understanding. The target outcome for the audience is "if I were in their shoes, I would do the same, thus it isn't so bad."
To further see the distinction, would the ethnic German fraction of the population explain other WW2-era German invasions: Poland, Benelux, France, etc? No. This was not an explanatory factor with predictive value. On the contrary, we know at the time it served as cover to a (false) hope that the Sudetenland aggression was a one-off.
With respect to the USSR->Russian Federation info graphic, the framing plays on the intuition that a something (country, capital city?) has a right to (control, oversight, governance?) of "its" citizens and economic production and that the post-USSR dissolution was a loss in roughly the same way that an individual losing property is a loss. I understand that the grey text explains and partially justifies, but that visual choice is pretty telling. The explanation text is extremely hard to read and is not what most of the audience will take away.
I agree. A better comparison would any other past empire wanting to invade countries that have since gained independence. E.g. the UK invading India, Kenya, or the US.
The UK doesn't border such regions. You could discuss the time period when English kings held part of Normandy and fought France for territory for a long time (the "hundred years war" for example).
Why is that significant? The original comparison ignored things such as culture, language, and how long these nations have been a part of the empire, so we might as well ignore proximity too.
I said taking into account culture was a good point https://www.astralcodexten.com/p/links-for-february-2026/comment/210410874 And sharing a border with a country obligates some kind of relationship (sometimes friendly, sometimes hostile, sometimes a mix) in a way that being an ocean away doesn't.
It borders Ireland
Northern Ireland does indeed border the Republic of Ireland, which is why the UK is more entangled with them. Ireland was not in the list of "India, Kenya, or the US".
Whose false hope? The british public's? Chamberlain went straight home and started rearming.
Yes, the British public.
I am definitely not a Chamberlain expert, but with respect to the PM himself, it seems there is a big range of reasonable interpretation. I endorse the view that Chamberlain was not a one-factor thinker who was overwhelmed by the idea that "he is only interested in ethnic Germans and will stop at the Sudetenland."
To me, however, "went straight home and started rearming" is misleading. It sounds like strong evidence that Chamberlain gave no meaningful weight to that argument, but I don't think it is supported by the actual defence spending stats.
The rearmament defense spending had been, arguably, underway for several years prior to the Munich Agreement. Data I found, link below, shows the following UK defence spending from 1934-39:
1934 120 mm
1935 120 mm
1936 150 mm
1937 200 mm
1938 210 mm
1939 270 mm
(the data is presented in GBP bn, which I am interpreting here in millions, but I hope this doesn't give the wrong impression about the level of precision of the data)
The increases from 1935 to 1937 were larger in absolute and relative amounts than in 1937 to 1938. Also, it might be useful to see these in context of the defence spending that suggest the 1938 budget of 210mm was not a particularly hawkish stance:
1920 690 mm
1921 300 mm
1922 220 mm
Qualitatively, the public addresses (like the Jan 1939 Birmingham speech) suggest that Chamberlain did have significant hope that German expansion would stop at the Sudetenland.
Links:
https://en.wikipedia.org/wiki/Neville_Chamberlain#Premiership_(1937%E2%80%931940)
https://www.ukpublicspending.co.uk/spending_chart_1920_1940UKb_17c1li011mcn_30f
The ethnic German population in Poland in 1939 was significant enough that, yes, you can explain that invasion as well based on ethnic reasons. Danzig was a majority German city, Western Poland had been under German rule until 1918. Southern Poland had been under Austrian rule and there were still plenty of Germans in Polish Silesia. Kattowitz was German enough that serious fighting took place when it was assigned to Poland by the allies.
I will go farther and say that the "they used to control it, so they are likely to invade it" argument does work for Poland (I am not sure if it works for all of the territory they took in 1939 or not.)
Between Sudetenland and Poland, though, was further occupation of Czechoslovak territory. That seems to falsify the claim that Germany was *just* restoring former territory and/or areas with high ethnic German concentrations. No?
The annexation of the Protectorate of Bohemia is indeed why Chamberlain decided Hitler wouldn't abide by any deal, so the next step (Poland) meant war.
In AJP Taylor's opinion the establishment of Protectorate was accidental and not planned by Germans.
I've slightly edited the text to say it's more about a general Russian view of geopolitics.
I think that was a mistake. There is nothing specific to Russia here. China claims Taiwan for the same reason. They also claim (unofficially) the borders of the Xing dynasty. It's perfectly normal for countries to claim past territories and territories with similar ethnicity. The US wasn't particularly thrilled to lose the south when it seceded either.
I was always irritated, then disgusted by Russian chauvinism (lived and worked in RF/UKR for 9 years between1994-2008). I would give my life if I could take Putin's. But, sure, an interesting map (not a territory!); thanks for sharing!
I don't "sympathize" with Putin, I think he's awful. The map conveys info for Americans who lack intuition about the size of such countries/regions within the former USSR.
Your analysis only makes sense if you assume that everything Hitler did has to be bad. It doesn't.
The idea itself come from this https://digital.library.cornell.edu/catalog/ss:19343225 which was a direct propaganda campaign against the Treaty of Trianon. I’d be 90% sure whomever made the map did so as pro-Russian-expansionism propaganda. That doesn’t mean it’s false, bad or unworthy of existence (Casablanca was propaganda) but I’d be cautious about passing it on uncritically.
Notably as well, they’ve done it in such a way as to make ugly janky borders, as opposed to modern Russia’s basically cromulent boundaries.
That map acts as if culture is meaningless. You can't compare New England to the Baltics, New England is arguably the cultural heart of America. Losing New England and Boston would be more like Moscow losing control of St Petersburg and Novgorod. A better comparison to the Baltics is probably Hawaii.
I will accept Belarus = Ohio because most Russians don't understand intuitively why Belarus is an independent state. Ukraine is trickier. Kiev is in some way the heart of "Russian" culture, but it is has never really been integrated into Muscovy, which is what the modern Russian state actually is. Texas is probably a much better comparison than California - the Lone Star state has a historical memory of not being part of the US, Texans have a strong local patriotism and a lot of Texans feel about the federal government the way Ukrainians feel about Moscow. Also Texas has played an outsized role in the way foreigners see the US, just as a lot of Ukrainian cultural baggage (cossacks, borshch) is associated by foreigners with Russia.
Almost no one in Russia has any nostalgia for bringing back the Stans or the Caucauses. I suppose you could posit that a rump US based around the historical 13 colonies and midwest might feel that way about Lousiana and Mississippi but really Puerto Rico and Guam are better cultural comparisons, even if economically less significant. There aren't even significant Russian minorities left in those places other than Kazakhstan. Most of the Russians in Georgia are recent arrivals.
This entire comment is points that I was about to make but Hannes explains them better than I would have.
Culture is a good point. The US is so young we don't even have the accent diversity over our wide geography that much smaller Britain has.
Britain had a uniquely large accent diversity, due to the extreme balkanization of small towns in Britain. France, in comparison, had a much less extreme version of "accent diversity" because towns regularly knocked knees with the next town over.
France used to have langue d'oc vs langue d'oil. Or maybe it still does.
I've never heard that before, why would British people have been so much less mobile than French?
Britain also has extremely strong class-based accent differences.
I don't know enough about enough countries to say what's normal and what's not, but I will note that Germany also has strong regional accents, and I think the vocabulary has more regional variation than in English as well.
Paranoia (this is the standard reason given for why English cuckoldry was often done within the village, as opposed to the French standard of "do it the next village over, as an affair")? Lack of long-distance rivers, which made travel much less profitable? England's civil wars might have also had something to do with this (France was more established as a "real country" for a long time before England actually managed to weld itself together).
"Class-based accents" are a later addition, coming from when Received Pronunciation was developed (unless you're speaking of Norman versus Viking accents, which was a whole different ball of wax).
> I've never heard that before, why would British people have been so much less mobile than French?
They wouldn't.
France, like pretty much everywhere else, made strong efforts to centralize the language spoken within its borders, which was significantly 𝗺𝗼𝗿𝗲 diverse than what we see in England. English in England is also much less diverse now than it used to be in the past, but it never achieved what the Romance languages did, because England is a tiny place.
As TGGP originally commented, there is more diversity of English in England than there is in the USA because English has a longer history in England.
At the time of the French Revolution, most French did not speak French, but a local patois. (*The Discovery of France* by Graham Robb)
It is said that in the alps, people in one valley couldn't understand the people in the next valley over, for both German and French dialects.
However, unlike English, German and French both have sort of central authorities defining what the language is, which I think probably contributed to the reduction of these differences.
Russia has hardly ANY accent diversity - including Russian speaking parts of Ukraine, where I lived some years (only one I heard of was: Is it ARboos or arBOOS - for melon), while I can not understand southern-US talk without subtitles. Just saying. Also, ofc: PTN PNX (ie: screw Putin)
This is a very good comment, and Texas == Ukraine gets to a lot of the heart of "Why Russians think Ukraine should be theirs" and why the Ukrainians do not (but might allow that Russia should be theirs, and controlled from Kiev).
(Puerto Rico has a lot of cross-cultural stuff with New York City, but that's mostly Puerto Rican immigrants, not the other way round).
>Almost no one in Russia has any nostalgia for bringing back the Stans or the Caucauses
Chechnya, Dagestan, and a few other obscure Caucasus republics are part of Russia now? Famously, Russia fought two very bloody wars to keep them. Did you mean Armenia & Azerbaijan instead?
The parts of the Caucases that lie outside the Russian Federation. You could compare Russian feeling towards Georgia to the way U.S. feels about Mexico. It’s definitely not America but we won’t tolerate it being allied with an enemy power, we like to vacation there and the food is not just a lot better, it has also become part of our culture. New Mexico is like Dagestan. It was Mexico, has a different culture but we aren’t letting it go. Sochi is like LA. It’s been firmly integrated and is now just Russia, the way Americans never think about how a town with a Spanish name is just part of the U.S. now.
#33 could have done a better job of matching the approximate cultural as well as demographic and economic relationships, but it does I think a pretty good job of showing how Vladimir Putin is basically channeling Londo Mollari at his worst (https://youtu.be/YbckvO7VYxk).
And it's an explanation, not a justifcation. Putin remembers Russia at the height of its imperial power and glory. He was there. He was part of it, and it defined him. Now it's gone, so no, he's not going to care whether it's "justified", he just wants it all back the way that it was.
The map, translates "the way that it was", into terms Americans can easily understand.
This is a poor characterization of Putin, and of Russia's stance in general. They are not pulling "Fortress Russia" out, as their playbook. (America on the other hand...)
"Above all, we should acknowledge that the collapse of the Soviet Union was a major geopolitical disaster of the century. As for the Russian nation, it became a genuine drama. Tens of millions of our co-citizens and co-patriots found themselves outside Russian territory. Moreover, the epidemic of disintegration infected Russia itself."
I also don't think it's a good way to build intuition, but not for the reasons you stated. I just don't get what intuition it's supposed to build. How much weaker Russia is compared to the USSR? But everybody knows the USSR was a superpower during the Cold War and Russia was barely a great power after the collapse. The relative size of Russia and the former Soviet republics? But I knew that already; there's a reason Russia is still a lot more powerful than Ukraine or Latvia.
Alice: "I'm going to kill George, steal his stuff, thus becoming a trillionaire!"
Bob: "That's terrible but Wow, unexpectedly the math checks out. George apparently has 999 billion in his Roth IRA and Alice has 2B herself."
Carol "Why are you trying to justify murder?"
I guess you could call Andrew Lee…the Joseon One.
Or Choseon one, it's sometimes Romanized with a Ch.
Mathematician David Bessis had a good article about the flaws or issues with most twin studies. Seems relevant to the “missing heritability” debate.
For #57. I think Threads is big in Asia. But I suspect most users are outside the US.
I recall Bessis being discussed in the comments here before, but then someone provided a link to what he'd written.
I can't find any relevant mentions of Bessis in past comments. Was that in a "for subscribers only" discussion?
Him and Jay Joseph were both mentioned. I dismissed Joseph as a crank (Eric Turkheimer called him "anti-science").
But it's clearly the TRUMP AMIERICA AI Act. Do they know how to spell down there?
I note the missing I was from "Intelligence", which seems in character.
Less important, but if you try to break it into a series of clauses or something the parallelism is super-jacked.
Like, I get:
The Republic-Unifying,
Meritocratic-Performance-Advancing,
Machine-Intelligence-Eliminating,
Regulatory-Interstate-Chaos-Across-American-Industry
Act
I have thoughts on the hydrofoils thing, drawing on a writeup I did of the concept a few months back: https://www.navalgazing.net/Exotic-Hulls-Part-3
I do not buy control systems as the reason for the Boeing Jetfoil's failure, because it is clearly and transparently false. Yes, in the mid-70s, computers were rare and expensive by consumer standards, but this was far less true in industrial applications. The F-16, which was fly-by-wire (similar technology for planes) first flew around this time, and Boeing kept delivering Jetfoils up through the 90s, when the program was sold to Kawasaki (because it wasn't profitable enough and most of the customers were in Japan.) They delivered more, which remain in use, including one delivered as recently as 2020, and another one is currently on order. If it was all "control is too hard", they could easily have replaced the system at some point in the last 30 years. They probably have. Also, not mentioning that they're still around or the reasonably long history of hovercraft ferries in other contexts makes me very suspicious of the startup's story. (Although this is a startup, so it's not like I'm surprised.)
The other thing that doesn't get mentioned is high maintenance costs. Yes, not needing engines will help some with that (and even engined hydrofoils will get better because they won't have to do silly things like pump water up into the hull where the engine can push on it, then back down to the outlet) but there's also stuff like foil erosion that makes them really expensive to run. Maybe they're going slower enough that's less of a problem, but I'd want to see some longevity tests first. The 6-passenger market is definitely a better spot for them than anything big (I suspect one of the big killers in the ferry market was that it can't take cars) but it will be interesting to see what comes of this.
Edit: As for naval use, I'm very skeptical. As always, my question is "OK, what systems exactly are you mounting on this, and what do they do?" Maybe there's a niche for inshore use, but I really don't expect a comeback for the military hydrofoil any time soon. There are options with much better track records that we aren't using.
Setting aside whatever issues there may be with the technology, I just don't get the economics of this. Apparently the largest version seats nine passengers. There are obviously cities where ferry service can be a fairly significant portion of public transportation, but the ferries are much, much bigger. If you want to run a whole lot of ferries, you eventually run into traffic issues, especially in terms of dock space. And you need good connections to other forms of mass transit at both ends (or you need to have docks that are within walking distance to major destinations). I guess there are niches where this could be useful, but most of the niches would probably be popular enough to be served by a bigger ferry.
Based on the article, they're talking about Uber/Waymo for water taxis. Which might not be completely crazy in the Bay Area if you can charge a premium for Uber-hydrofoil-Uber over just doing Uber directly, particularly at rush hour. But yeah, there's going to be scaling concerns, and while the obvious solution there is to make a bigger hydrofoil, we've sort of had those for a long time...
Navier's actual marketing seems to be to people who want a hydrofoil boat.
I guess it works if you can guarantee that the wait times are minimal for every leg of the trip. If you have to wait ten minutes for the second and third legs of the trip, you're probably eating into a lot of the time savings. And this boat is going to cost a lot more than a car, so you need to charge a much higher price for an equivalent return on investment.
The transit consultant Jarrett Walker wrote a good post a while back about what characteristics are good for ferries to succeed: https://humantransit.org/2016/12/ferries-opportunities-and-challenges.html
There are a few cases where they work well (Staten Island and Hong Kong, as well as the river in Brisbane) but they have an inherent limitation in that every station has no development possibilities on one side (being at the water) and the ferry terminal itself needs to be at some of the most expensive land in the city (waterfront).
Another issue is that many of the places where all these criteria are met are busy waterways which have modest speed limits anyway, so the hydrofoil is useless.
Do you really need a terminal? A dock seems like enough.
NYC Ferry terminals produce nearly no obstruction on the land. They consist of a small fare/waiting area and a pier.
I agree about the bottleneck being something other than the control systems, but I do not think the high maintenance costs are unsurmountable either. For example, Russia is using a large fleet of Raketa/Meteor/Kometa hydrofoils as water buses on its lakes and rivers since late 1960s. They are passenger only, but seat between 60 and 120 passengers, so are not quite at the taxi scale.
I am well aware of those, having done a deep dive into the uses of the hydrofoil about 6 months ago, but I'm not sure they prove your point. Those were produced in the Soviet Union, and cost-effectiveness was not always high on their list of priorities. Now, on the other hand, they did have some export success, but I don't remember how long those tended to stay in service after being sold abroad. And since the 90s, it could easily be "well, this is what we have, keep it working".
The other issue is that I don't really understand the foil erosion problem that well. It was definitely an issue for the Pegasus class, which I have by far the best information on. But how can it be mitigated? Was it mostly because the Pegasi were so fast, and cutting speed makes it go away? Was it because the Pegasus was submerged-foil and the control system couldn't handle the loss of lift from the erosion? Or just that it had no margin to handle the loss of lift when trying to get up on the foils? A smaller surface-piercing foil at lower speed addresses all of these, but I'm not sure that the real answer is even on my list.
Modernised versions are still produced and used in Russia today, e.g. Meteor-120R, launched in 2022. Cost effectiveness was not much of a concern in Soviet times, but it is as much a concern to current Russian operators as it is to anybody else. In Estonia and Lithuania, Raketa produced in 1960-1970s stayed in service until early 2000s (https://en.wikipedia.org/wiki/Raketa_(hydrofoil))
As far as I can tell, it's just Meteor-120R, and that seems in very limited production. And even there, I could easily believe that in a few cases, it's cheaper to build a new hydrofoil ferry than to redo a city's transportation infrastructure to use a different method. Which is not the same as saying it would have made sense from the start. But this whole thing is deeply mysterious to me, and will remain so unless someone runs across a careful academic analysis of the operating costs of various high-speed ferry options. I'm not looking, because that series is done (ish) and I have other things to write about.
57. I use Threads. For some reason Threads became popular as a Twitter alternative in Finland (perhaps it was simply launched here at the same time as there was one of the Twitter exoduses?) and the community is probably currently better there than on Twitter, so I've moved there rather organically as well. There are probably some other nations where this has happened as well, just like Twitter communities of some other nations have moved en masse to BlueSky, Mastodon etc.
The Finnish Threads community is basically stereotypized as "horny late millennial white collar workers", ie. what Twitter used to have before the... acquisition. In my experience the American Threads community seems to be heavily African-American, discussions on topics related to this community pop up on my timeline all the time.
It's fairly popular in Taiwan as well.
#43 - Small fix: the country should be "Antigua and Barbuda" (not "Barbados").
Thanks, I never realized those were different places before.
My first thought reading this was "that's an insane thing to not know for an educated adult American" but then I remembered that you're a West Coaster who doesn't browse the Caribbean for quick vacation getaways and also that I live in NYC, with a very large Caribbean population.
Whose name literally reads "ancient lady with a beard" in Spanish.
On #57, I think it is just surprisingly easy to be unaware of the social networks "other" people use. If you use Twitter it is probably because most of your network uses it, and things like that tend to be sticky. Personally, I almost never use Twitter (and have been that way since way before Musk took over), know very few people who talk about using it, and have no real awareness of how many people actually do use it.
Twitter has a lot of mindshare with journalists, which has always exaggerated its importance.
Regarding murders and trauma care, what percentage of murdered people used to die in (or on the way to) a hospital?
Improved trauma care only helps if you receive trauma care. If back in 1980 only 10% of murdered people died after an ambulance arrived then care could become instantly life saving and still barely impact the murder rate.
Seems like a potentially debate resolving question.
I had the same thought. In some cases, "murder" is essentially an assault attempt whose consequences exceeded the perpetrator's intentions. But more often, I think, murder is the result of the perpetrator deliberately killing their victim, and they'll stop when they feel adequately assured that their victim is dead.
Some types of assault, like knife attacks, are more survivable than many people suppose, and you can survive being stabbed once, or several times, with high likelihood, but that was already true decades ago, and people committing deliberate knife murder will usually frantically stab their target over and over until they stop resisting.
I don't think most murders are "intentional outcome" murders. I think most murders are part of dispute resolution, in cases where cops/law enforcement cannot be called. And there, folks ain't trying to kill the other guy (as that gets his gang madder'n hornets)... You get "heat of the moment" reactions, and that leads to "calling 911", for damage control if nothing else.
Do you have a source on this? I've never been in a gang, but I've lived in high murder rate/ high gang activity areas, and talked to people who've been in them (in some cases,. maybe still are?) And this doesn't correspond to the experiences they've related, at least.
Not that sub-murder gang violence doesn't occur, but I've heard a fair number of firsthand accounts of violence, and all (except possibly one uncertain case) which led to death were intended to.
Sources yes. There was an incidence of "gangland violence" where "bring back a bloody knife" was the ask, in order to join the gang. Apparently, ganging up on the local dingus* turned out to be the actual "what happened." He was hospitalized, as there were more than one kid who wanted to join the gang.
You have "beefs" that get "resolved" at neutral places, like bars. You'll have "violence resolution specialists" that provide counseling on "how not to draw guns when drunk" and other things like that. This wouldn't be a "big thing" if there weren't deaths/bad injuries to prevent (https://scholarworks.lib.csusb.edu/etd-project/3209/)
Yes, there are actual "murder this guy" jobs, and people do have actual jobs as assassins as well, sometimes under the employ of state actors. Those are relatively rare, compared to "everyone who has a job in the gang."
There are a few other things going on -- you may not be getting the whole picture, on "trying to kill someone" (there's incentives to say "yeah, I totally meant to kill him" as it makes you sound badass). Said critique applies to my sources as well, "bleeding hearts" may do interventions that save only a small fraction of kids, simply because "one kid saved" is a net good.
And I've discussed with a source about LA gangs calling 911 (we're not talking about ethnic cleansing, which would obviously not involve calling 911, as the objective is murder).
*when a little guy asks the drunk, "are you afraid of me" we aren't exactly surprised when he says "no" -- but when the little guy draws a knife, the correct response is not to continue insisting no, but to RUN.
This is all largely familiar to me, but also all seems consistent with "most assault attempts and instances of violence do not lead to death, most deliberate murder attempts succeed (or at least are not prevented via quality of medical care.)"
Your account of gang life is fascinating and I want to thank you for sharing it, but aren't most murders committed in the context of a domestic violence situation? In which case violent crimes and assaults could be going down because as a society we have learned how to handle domestic conflict more effectively.
I'm not sure. You have business, and then you have "reputational business" (which could be DV if you stretch it -- shoot the woman at the bar, because she's disrespecting you by flirting with another man -- or shoot the man, because he ought to know better too).
I'm not sure in a country with a 3x suicide to murder rate that we know where most of the dead bodies are. And that's not getting into the "excess Deaths."
Don't fall for the frame until you've investigated all the "easy avenues" to call something "not a murder."
Relevant (old) article. Seems GSW mortality flat for this reason:
https://www.thetrace.org/newsletter/analysis-shooting-victims-gun-violence-intensity-trauma-baltimore/
While reliable statistics are hard to come by, as is often the case with anything involving guns,* it appears that about 80% of shooting victims survive. That’s for handguns; rifles and shotguns have much lower survival rates, but are seldom used in crime. I do not know whether the 80% figure includes self-inflicted shootings.
* = Undoubtedly for political reasons no agency keeps records on how many justifiable defensive shootings by non-police occur each year.
Answering my own question now that I had a chance to look up some data:
EMS only responded to ~70% of homicide victims in 2022. (https://www.cdc.gov/mmwr/volumes/74/ss/ss7405a1.htm)
2007 review of patients of gunshot and stabbing victims being brought to Philly level 1 and 2 trauma centers found 16% were DOA and another 11% died of their wounds at the hospital. (https://penntoday.upenn.edu/news/survival-rates-similar-gunshot-stabbing-victims-whether-brought-hospital-police-or-ems-penn-med)
Another study showed gunshot wound mortality going from 16% to 10% among admitted patients over the course of 20 years (https://pubmed.ncbi.nlm.nih.gov/28697020).
Back of the envelope math: 30% do not get an emergency response. 41% get a response but die before reaching the hospital (70% x 16/27). 29% get a response and die at the hospital (70% x 11/27).
Dropping mortality from 16% to 10% on admitted patients only reduces deaths for the 29% that arrive alive. That would be an 11% overall reduction in murders from improved care.
Take away: I put my estimate on murder reduction due to increased trauma care at 20%, with a range of 10-25%. My data was from different places at different times but is consistent with other articles that I read over in my admittedly quick review.
A simple extrapolation of 11% over 20 years is 27.5% over 50 years, pretty close to the claim that "murder and other violent crimes have “only” gone down by a third in the past fifty years".
Linearly extrapolating the gunshot mortality rate going from 16% to 10% is probably wrong, especially if you're extrapolating that far (30 years). If we extrapolate 34 years past the end of the study, the mortality rate will be negative.
"Blessed is the mind too small for doubt [in straight lines]"
EMS provide some emergency care on site and while in transit and I would expect the outcomes of this care to improve alongside hospital outcomes improvement. So just counting the 16% to 10% improvement in post-admission outcomes underestimates the overall improvement of victim mortality from medical advances.
Can you explain why you did 70% x 16/27? Shouldn't the percent who get a response but die before reaching the hospital be just 70% x 16% = 11.2% ?
It is because one is a subset of the other.
Of all homicides, 70% had an EMS response. Of all EMS responses to gunshot/stabbing victims 27% died (16% before hospital, 11% at hospital). Because we are only looking at deaths, that 27% death rate represents the entirety of the 70% with of murder victims who had an emergency response. I applied the ratio of where people died after an EMS response (16/27th before hospital, 11/27th at hospital) to the 70% of murder victims that had an EMS response.
Robbery-murders seem way less common today than in, say, 1980. I seldom hear about 7-11 clerks or mugging victims in Central Park being shot by armed robbers compared to a couple of generations ago. Armed robbery seems down in general, perhaps because there's less cash payments compared to electronic payments.
Our society is better set up to discourage rational crime than it was 60 years ago when crime was just beginning to rise, so most murders these days are due to stupid people getting in stupid beefs with each other.
Trauma care won't help with intentional murder - if the goal is killing, the perpetrator will indeed shoot the corpse. It should reduce manslaughter incidence - especially vehicular manslaughter. I have no knowledge of what technical crimes are included in "murder rate down" statistics.
Vehicular manslaughter at speed tends to lead to people with enough trauma that EMTs say "died instantly." This is nearly always a lie, but did you really want to tell the grieving widow that her husband died screaming as he burned alive?
On the murder survival rate, thing; the increase in survival for assault victims is real... but it mostly happened decades ago (70s, really); there has been little change in the last 40 years or so. You'll often see people quoting assault survival rates that exclude people found dead before reaching medical care, which rather muddies the waters.
Recent trends have actually been for *worse* trauma survival rates, not better.
> chomping at the bit
It's "champing".
As a fellow pedant, I stopped correcting people on the champing/chomping thing when I realized the latter is just an American regional variant of the former. I’d never correct someone for saying “I don’t give a rat’s arse”, so why bother with the bit chomping 🤷♂️
Even absent the regional meaning, "chomping" is a pretty good description of a horse biting a metal bit! It's an intelligible phrase in its own right and has a nearly identical meaning, like saying "throw the babe out with the bathwater" instead of baby, or "put him on a plinth" instead of pedestal.
It's not good style, obviously, but it shows a familiarity with language and metaphor that's absent in phrases like "could care less". It's arguably more consistent with the rest of the language, too.
Interesting that I saw how coding LLM did 1+1 calculation without any reason during thinking. It was in JetBrains IDE chat interface. I am puzzled because this was not a tool, it was just directly during reasoning. But it might be related to the calculator tool example somehow.
It's probably not serious anyway, but your dating startup would instantly fail due to nearly all men giving a 9 to nearly all women, and vice versa nearly all women would give 0 to nearly all men. See swiping rates per gender.
I would imagine it would scare away most people because rating from 0 to 9 is too complicated, and those who remained would use it more as intended.
Related: OKCupid, before it became a Tinder clone with fake accounts, was actually a very good dating site because it scared away people not willing to write a good profile and answer questions (at least for me; I got more dates out of it than from any other site, an most of them were at least interesting).
I'm aware of OKCupid, but that appears to me vastly more complicated than just a 0-9 rating. It's a possibility, though.
If users aren't encouraged to write good profiles or include much in-depth information about themselves, I wouldn't expect the 0-9 rating to be very meaningful as a predictor of compatibility anyway.
All systems need ways to ration scarce resources/awards. It is weird how often systems don't do that and the system falls apart, see grade inflation.
I am just interested in the psychology of people who don't want a rationing system.
Correct, though what's the best rationing system here? Intuitively one may ration by high scores, but arguably, both very low and very high scores should be rationed.
> Correct, though what's the best rationing system here? Intuitively one may ration by high scores, but arguably, both very low and very high scores should be rationed.
I think on the dating app side, it's already pretty good - there's apps like Raya and The League that have entry barriers (at least on the male side, I think on the female side you just have to be hot, basically), and sorting by quality at the app stage is basically what you need for a stable matching problem to work.
The big problem is that the non-selected dating apps don't work for below top-quintile men and women to actually pair up in LTR's.
And for that we need something else entirely - we need matchmaking.
All the matchmaking things in the US kind of suck - tiny user populations, high costs (certainly higher than 80% of people will pay), and so on.
So what we really need is a cheaper, high quality matchmaking thing that can apply to the 80%. Personally, I'm hanging my hat on AI.
When everyone has AI personal assistants in their ears that know everything about them, it should be pretty easy for the AI's to help match people up, in a way that's much higher fidelity and better compatibility than the people themselves manage.
I was thinking more in terms of Scott's specific proposed dating app, less on the general dating app market. Otherwise I have little to add here, since I have almost no experience with contemporary dating apps and no inclination to change that.
On the AI personal assistants, I increasingly fear that the same people currently struggling to find a partner will just forgo finding the irl romantic partner and just pretend that the AI personal assistant is one.
> On the AI personal assistants, I increasingly fear that the same people currently struggling to find a partner will just forgo finding the irl romantic partner and just pretend that the AI personal assistant is one.
Yeah, but to be honest, that sounds like winning to me, too?
Literally every generation of women since the 50's has had lower marriage rates:
https://imgur.com/cV2Sbdv
This is largely a good thing, because the vast majority of people suck. When ~90% of women were married in the 50's, it was an immense charnel house of suffering and wasted potential.
How do we know this? Even *today* divorce base rates are ~42% for first marriages, and about half of the remaining marriages are net miserable for at least one party. That's a roughly 2/3 failure + misery rate. Think how net miserable those 90% of marriages were, that could NOT get divorced, back in the 50's!
As soon as no fault divorce opened up in the late 60's, divorces surged and that was a good thing - it was millions of soul-crushing relationships suddenly able to be dissolved! Not to mention that all the actual female talent that had been shackled to soul-crushing duds were now free to get careers and open bank accounts and live more meaningful lives, which was also great for GDP.
And likewise, the fact that each subsequent generation of women has married at lower rates is a great thing! That's millions of shitty relationships, not being entered into! It is very much a "that which should be destroyed by the truth, should be" sort of scenario, and we should rejoice that it's happening.
So when this trend is accelerated by AI, who will be superhuman at relationships in basically every respect (at least when whatever soon-to-be trillion dollar company gets good sexbots worked out), that TOO will be a good thing, for all the same reasons.
Fertility crisis? We'll have ~8B people through year 2100. Serious declines don't hit for another 100 years. 200 years is more than enough for AI technorapture or transgenic pigs or some other artificial womb solution to get worked out.
>Fertility crisis ...
Humans aren't fungible. Here in germany, the age pyramid has already gotten so lopsided that it's entirely impossible to actually realize the retirement promises that the state has given to the public all their life, and this will most likely get worse in the 2030s. Immigration exacerbated this issue, instead of improving it as hoped. In the longterm evolution will sort things out, as always, so I'm not particularly worried, it's the near- to medium-term that looks bad. Depends on the country though, the US seems relatively fine.
>Divorces ...
I guess I just simply do not buy into the entire framework. I grew up in a conservative christian region, with marriage & divorce rates more similar to the 50s then anything else (though the legal environment was pretty much modern already). I became an atheist and moved to a liberal university, where most of my new friends unsurprisingly came from a liberal background, and now that I'm older I also get to see urban liberal coupling patterns up close.
And the more I see, the worse it looks. The higher percentage of people who break up aren't happier; They just falsely blame their problems on their partner, falsely believed that a single life and career/self-actualisation makes them happier or falsely thought they could "do better". I'm aware of surveys on the topic, but I have absolutely zero confidence in their reliability; Nobody has seemed more miserable to me than a childless single women in her 40s proclaiming how much she enjoys her freedom. And they do occasionally admit that they regret their decisions, though rarely publicly, and rarely for long.
The kids also suffer a lot under a divorce, and making family life nice simply isn't just about the right match (though it certainly helps); It's about adapting to your partner/kids and putting in the work. About priorities and just letting go of your individuality (not entirely, of course, but partially). As far as I can judge it, liberal reddit-tier "always focus on your needs and break up over every inconvenience, you go grrrl pwr" is the worst thing that has happened to human happiness in a long while. And this even partially applies to liberals who do stay together, because they regardless just can't help but always blame their partner, always dream about how much better their life could be without any obligations or with someone "better". And it's also part of a bigger package that means well but just makes everyone unhappy, such as therapy and guilt culture.
The conservatives aren't perfect, of course, and their divorce rate isn't zero, either (which is good - I'm not against divorce in general, I just think liberal culture is far too biased in favour of always breaking up). But even my wife, who comes from a more liberal background and had lots of ultra-liberal friends in her early adulthood, just can't shut up about how nice, happy and well-adjusted everyone is when she meets my family or my old school friends. I know lots of liberals are really into the idea that it's all some stepford smiler hellscape, and some people like that surely exist, but no, you actually can just be happy with the people you have instead of making yourself and others miserable by always trying to get something better.
But I'm aware this is something that probably can't convince anyone who hasn't personally experienced the difference.
Rating should be on a curve, if you give everyone a 9 then that should be counted as a 5.
Encouraging honest rankings is one of the problems for voting systems in Arrow's Impossibility Theorem.
Well... sort of. One of the mutually-incompatible criteria is monotonicity: ranking a candidate higher in your personal order of preference should have a nonnegative effect on the electoral outcome of that candidate.
But this isn't a problem for the system you're talking about, where you rate people on a scale from 0 to 9. Arrow's Impossibility Theorem doesn't apply to that at all, because you're rating people in cardinal numbers.
An ordinal ranking of people WOULD obviate the problem of people just using the maximum/minimum numbers available.
Give people a limited rate of points.
(Also, would they? I definitely got women on Papa's sometimes who I'd give a 4-5 under this system)
Dammit did we just reinvent onlyfans from first principles
He’s mentioned in the past that this is designed for people you know in real life, and aren’t passionate about but would be sort of interested in. I don’t think normal swiping rates per gender would apply to real acquaintances.
That's a whole 'nother can of worms - I remember an app that would allow you to "secretly" check a box for facebook friends you'd be willing to date, and they would only see this if they *also* get the app and *also* check that box. Seems simple and uncontroversial, right?
Except when some women heard about the existence of it, they'd get the app, check everyone's box, and then post publicly about those disgusting creeps who checked hers. Dating among irl acquaintances really needs plausible deniability and/or very well-honed social instincts, especially in modern culture.
> Except when some women heard about the existence of it, they'd get the app, check everyone's box, and then post publicly about those disgusting creeps who checked hers. Dating among irl acquaintances really needs plausible deniability and/or very well-honed social instincts, especially in modern culture.
Did this really happen? Seems obvious that some will exploit the system in that way to gather information, and sadly probably inevitable that some will make it public. But did any actually frame it as "disgusting creep" rather than (also bad, but at least not moralised) "lol, loser"?
Which is why it would be simpler to collapse it into yes/no and match on yes/yes. Wasn't this used in at least one previous dating app?
What if you could only get ten potential matches each week and had to rank them, not rate them?
"Can’t believe he missed his chance to make Georgia Georgia." 🤣
The link to the Claude chat on cancer and dementia seems to have a permission issue.
Perhaps you took your own advice from the AI safety item?
I can see the chat, but when I got to the question about terminal lucidity the response was only half generated. Scott, did you get a full answer that's just not coming up for me?
When I try to view the linked article "Barsoom - Amelia Sans Merci", about a British government-sponsored project to discourage people from researching right-wing ideas online ... I get a message telling me this is "Age-restricted content" and I must verify my age to continue. Substack helpfully explains that this is required to comply with British government regulation (the "Online Safety Act").
In other words, in order to view subversive content, I have to identify myself. This seems like it might cause a chilling effect.
Yeah, the Online Safety Act is ridiculous.
https://rairfoundation.substack.com/p/uk-government-backed-anti-extremist
https://www.americanthinker.com/blog/2026/02/all_about_the_u_k_purple_haired_conservative_internet_sensation_amelia_the_victorious.html
Substack to the rescue!
We thought that opening up trade with China would make them become more like us. Instead we became more like them.
<quote>When $100,000 is the limit available for an insurance claim, 96% of personal auto claims settle below $100,000</quote>
What is this supposed to show? Isn't the point of insurance to limit worst case risk, not an average case?
In the case of auto-accident insurance, there's a perverse incentive for claimed damages to adjust up to or just below the defendant's insurance limit, because the defendant will probably just settle at that level if it makes everything go away (modulo higher insurance rates going forward), but will fight if the alternative is a big chunk of their life's savings.
Dialing your personal liability insurance up to $1E6 rather than $1E5, has a very small possibility of protecting your assets in the case where you e.g. really did inflict $1E6 of medical bills on someone and they have the receipts and really need the money. But in the much more likely case where you are involved in an accident and the other guy's medical bills are say $20,000, the answer to the question "now how much 'pain and suffering' should I ask for on top of that?" will tend to be $980,000 rather than $80,000.
Demanding that *other* people carry $1E6 in liability insurance, looks a lot like a gift to plaintiffs and their lawyers.
So how do we insure against when someone "really did inflict $1E6 of medical bills on someone"?
I can't think of an answer that doesn't involve large-scale tort reform.
My first guess would be a version of "loser pays" where it is presumed reasonable for a defendant to incur legal expenses in their defense up to 15% of the claimed damages, and if the plaintiff's claims are (mostly) rejected by the jury then they and/or their lawyers are on the hook for that. That would encourage the people with $20K in documented medical bills (etc) to ask for that and maybe a little bit more, but threaten a big downside risk if their larger claim is rejected. And note that in the first case, the defendant('s insurance company) will be paying out not just the $20,000 or whatever, but if they don't settle will also have to cover $3000 of the plaintiff's legal fees.
But there are lots of second-order effects and unintended consequences for anything like this, and I haven't put a lot of thought into it.
The US is something of an outlier in how often and how large non-economic and punitive damages can be. Other countries restrict this more.
US law could have restricted the liability to medical expenses and direct economic damages.
Or put bounds on those costs, which I understand some states have done though even those limits can seem ridiculously high.
The victim can (and mostly does) get and use his own health insurance. It may be unfair to not have the party at fault pay for the costs, but it may be the lesser evil.
It's not clear to me from the linked Substack post whether we're talking about insurance for liability towards third parties (other drivers, pedestrians), or towards the passenger, or both. If it's about liability towards third parties, liability insurance is already required for all cars on public roads in most jurisdictions; there's no reason for the rules to be different between rideshares, taxis and private cars (but it seems like rules *are* different, with much higher requirements for rideshares).
If it's about liability towards the passenger, I see no reason to require liability insurance at all, or even have liability towards the passenger in the event of an accident. The passenger should be allowed to waive it even if the service provider would be liable by default.
After all, if you crash your own car, you either pay your medical costs out of pocket, or your health insurance pays them, or the liability insurance of the other driver does. When you use a rideshare, there's little reason to effectively require you to indirectly pay for extra insurance compared to that (via the driver's liability insurance getting factored into the price). In fact, if you already have health insurance, it sounds like you're paying twice to insure the same risk.
Theres also an opposing perverse incentive, where theres no point in fighting for more than the defendant+insurance can realistically pay, even if the damage is higher - and at 1E5, the additional amount you can extract from the defendant alone is often going to be small in comparison, and worth giving up for even modest increases in the odds of getting the payout. Note also that *this* effect is unavoidable for those harmed by it, while yours could be avoided by better control between the defendant and his insurance. The entire point of mandatory driver insurance was to avoid this problem.
I think the US does often have excessive damages, but the figure cited here is no evidence of it - how you interpret it entirely depends on what you already believe.
54. Note that those charts start around 1979, which was a local peak in criminality, creating a narrative of a consistent long-term decline in crime. It looks different if you start earlier and see the massive boom from from 1960 to 1980:
https://www.researchgate.net/profile/Angela-Dills/publication/5188722/figure/fig1/AS:668912853213185@1536492660055/olent-and-Property-Crime-Rates-per-100-000-1932-2006.png
Though you could raise the same objection to starting at 1960, crime rates were higher before that, as can be seen in the homicide rate graph:
https://static01.nyt.com/images/2026/01/21/multimedia/2026-01-07-crime-stats-index/2026-01-07-crime-stats-index-facebookJumbo-v4.png
Interesting. Thanks for sharing the second graph. The late Kevin Drum hypothesized that removing ethyl lead from gasoline was the cause of the crime drop. I tended to prefer that hypothesis, but that graph would seem to undermine that theory. The moral of that story is always zoom out (on graphs).
Yeah seems like all discussions about crimes is muddied by something else that's VERY VERY BIG. What is it? Lead? Communism?
Typo: "This problem may one day be solved by frontier labs may one day be able to change this" looks like two drafts of the same sentence accidentally spliced together, "may be solved by frontier labs" and "frontier labs may be able to".
9. I think I'm with DC here. If the argument is that settlements are too high, address the settlements. Otherwise it's a subsidy from victims of high-impact crashes, and the drivers/riders aren't paying for their externalities. It would be cheaper if we didn't require any insurance, but victims would be double-losers.
What does "address the settlements" look like, though?
Legislation, probably? $100k seems really low. If someone's crippled by an irresponsible driver, that doesn't go far. Just arbitrarily reallocates costs.
What would the legislation say?
To replicate this outcome consistently it would cap all damages at $100k. If that sounds crazy, then so should this. I don't know enough about law to give details about how it could work better.
Yeah. Under insurance limiting liability for bad car crashes is a real issue (effectively it means taxpayers or victims subsidize bad drivers).
TIL'ed weaboo/weeb is a word which should be odd given I live in a Far East Asian majority state. I feel this word has way less penetration that it seems; or I live in a bigger insular bubble than I thought lol.
The last part of #49 made me smile but honestly a bit much there, Musk may be a lot of things but a Nazi isn't one of them, not even neo. Sure he would have been had he been alive in Germany back then but so would most people.
I'm guessing people are less likely to fetishize Japanese culture when there are a lot of actual Japanese(-American) people around!
The reverse, idolizing British culture, is called being a "tea-a-boo".
And people who idolise American culture are called "freeaboos" :-)
Most murders aren't deliberate attempts to kill someone, it is usually using extreme violence to exert something and not caring if the person dies. Which morally and legally is seen as murder.
Gang killings are usually about exerting power in a dispute rather than deliberate assassination.
A murder is by definition a deliberate attempt to kill someone. Otherwise it is manslaughter. You seem to thinking of voluntary manslaughter where the perpetrator kills in the heat of the moment, using extreme violence without necessarily thining of the consequences, but not pre-meditated.
That is not the Common Law definition of murder.
It depends on where you live. In California, second degree murder can refer to intentional but not premeditated killing, or killing someone with reckless disregard for human life.
They're saying most murders are the second one. The perpetrator isn't specifically trying to kill someone and ready to make sure they're dead. They're just inflicting general violence, not caring all that much whether or not someone dies.
#55: Car theft is another reliable crime statistic, for no matter how much they may dislike the police/think them useless, owners of stolen cars nevertheless have to report the thefts in order to cancel the registrations and, in many cases, file insurance claims.
Also, while better medical care may be saving some crime victims, the higher capacity of today’s semiautomatic pistols vs. six-shooter revolvers almost certainly means more multiple hits.
I'm skeptical that we've seen changes downstream of better guns in the last 30-40 years. The high-capacity semi-auto has been the generic "this is what you get if you want a pistol and don't have other preferences" since the mid-80s, and while there's some lag in what ends up in crime, it's not that long.
3x more people in the US are killed by knives than rifles. More murderers beat their victim to death with their bare hands than use a rifle. https://www.statista.com/statistics/195325/murder-victims-in-the-us-by-weapon-used/
Rifles will never come close to handguns as a criminal weapon for the simple fact that you can't conceal a rifle.
Yes, and that might have been relevant if I'd been talking about rifles instead of pistols. I was instead discussing the shift from a mix of revolvers and single-stack semi-auto pistols like the Colt 45 (7 round mag) to double-stack semi-auto pistols like everything on the market today.
That's a fair point then. I saw "high-capacity semi-auto" and thought that referred to a rifle and that you meant people were buying those instead of a generic pistol.
Until the '90s US crime guns were largely shit like 6-shot pot-metal .25s, Rohm revolvers, and so on.
I do not believe that this is the case.
There's not going to be any statistics on the literal subject of "guns used in crime" because most crimes are e.g. muggings where nobody gets shot, nobody gets arrested, and the victim just says "he had a gun and it was bloody scary and I didn't get a chance to measure the damn thing". So you can imagine that all the muggers who aren't shooting anyone are using the crappy little Saturday Night Specials and I can't prove you wrong.
But we do have much better records on guns used in homicides over time, e.g.
https://forum.cartridgecollectors.org/t/cartridges-used-in-crimes/8323
https://www.ojp.gov/ncjrs/virtual-library/abstracts/homicide-handguns-and-crime-gun-hypothesis-firearms-used-fatal
Through the 1980s, the most common handgun used in homicides was a .38 revolver. And most of the rest were at least .32 caliber. From the 1990s, it's been mostly 9mm.
For historical context, the 9x19mm cartridge was invented in Germany in 1908 and gradually became the world standard over the next century, but it started in Europe and took a good long while to become popular in the United States. The so-called "wondernine" pistols that incorporated all of the features various European gunsmiths had been coming up with, came on the market in the 1970s and a few weirdos in the US thought they had potential but mostly people stuck with their .38 revolvers and .45 automatics. Mid-1980s, a number of major US police departments started using them and perhaps more importantly they started showing up in action movies in a big way, so after a few years of pipeline effects they became dominant.
That's correct about car theft, and applies to carjackings as well. In both instances the criminals are strongly incentivized towards taking newer/better/more expensive car models, and in turn the owners of same are more likely to have full-on car insurance coverage that is worth reporting a claim to.
The scenario of teenagers or gang members snatching a random beater just to have some yucks driving it around (a.k.a. "joyriding") does exist, but is statistically trivial.
There is a form of progress that cuts down on car theft over the decades graphed above: lojack.
Surely standard engine immobilisers are a huge factor as well? A 1980s car can just be hot-wired, a modern car will refuse to start without electronic contact with its own key.
Good point, although I believe there are a couple models of car where people were able to defeat that, spread the technique on TikTok, and then there were a spree of thefts of those cars. The manufacturers even had to distribute clubs to owners of that model.
Can you please post a link for the clubs thing? That sounds made up (if the owner sees the theft, there are probably better options). If it is made up, I don’t want to add to the search traffic for it.
https://www.elogps.com/kia-hyundai-settle-what-does-this-mean/
> 33: Interesting as a way to build intuition for why Russia invaded Ukraine, h/t @MMJukic
(i got carried away and wrote a more general point)
I think that the war has made everyone slightly more interested in Russian politics, and I worry that people from certain backgrounds (e. g. Americans) may participate in the discussion with certain assumptions they don't even notice, that don't actually apply well to the Russian political system, and which they have a difficult time discarding.
E. g.:
1. I suspect it's very hard for someone from a more distributed political system to internalize the degree to which a country's political system can be *purposefully, cleverly, resourcefully optimized* to tune out any feedback signal, so that in its decisionmaking, it can fully follow its own preferences, instead of having to listen to annoying outside signals (like the country's own population, independent politicians etc.)
2. There's this myth of The People finally rising up against the tyrannical elite; which ignores that (if understand correctly) revolutions are mostly about whoever is currently in power already being very vulnerable to being overthrown or straight up agreeing to leave of their own will, e. g.:
— Nicholas II abdicated in favor of his brother, who then also abdicated; bolsheviks didn't overthrow the czar, they only overthrew the extremely weak interim government;
— Ceausescu signaled his weakness with that disastrous televised speech, lost support of the army after the death of the minister of defense, and was overthrown by his own fellows who already played a role in the Romanian political system;
— the Armenian revolution of 2018 involved mass protests, but it succeeded when Sargsyan left the office of his own accord, releasing Pashinyan from jail where he had been confined since the previous day.
In other words, if you have no ways of coordination because your institutions have been non-functional for a while, if you reasonably mistrust your peers (who knows if they'll report you to the secret police), and if your regime is united, coordinated, possesses many resources, has an army etc., the regime might get away with many things everyone hates and without facing any romantic glorious rebellion.
(1 and 2 are also are both present in the sentiment one sometimes hears, "most Russians must be supportive of Putin's policy, if they weren't, they'd overthrow him". I don't actually know if "most Russians" support Putin or not, it's really hard to measure when speaking out is dangerous; it's just not a locally valid logical step.)
3. the idea of Putin's role representing his country's interests in *some* way, e. g. acquisition of resources; while I feel it's more accurate to model his decisionmaking as trying to keep himself safe and keep his job. trying to keep some form of popularity is important to this, utter military failure signals weakness to elites (and signals are something to coordinate around! they create common knowledge in the form of "everyone knows that everyone knows that <...> Putin is weaker than usual right now"), while a "short victorious war" signals strength, etc.
I think there are more examples of this, but I'm not the best person to describe this; I'm not competent at political science.
I think that #33 may *somewhat* suffer from point 3 (but in general, they all sort of overlap.)
It gives the relevant vibe (that I think how Russian political elite — Putin and like the security council, whatever) that there are resources for which countries compete, and its peoples are an annoyance rather than the what countries are made of.
But also resources are not an end to itself: you need to control more in this world to be able to listen to less people inside and outside, so that no one can inconvenience you in any way when you make decisions they don't like.
Nicholas II was indeed overthrown by a popular revolution, assisted by mutinous army troops in St Petersburg. That is what is known as the "February Revolution" in Russia. The Bolshevik "October Revolution" was really a putsch and a takeover of the existing revolutionary government, as you say not really a popular revolution despite the Soviet propaganda.
Maybe what I said reads like "he decided to abdicate out of the blue", which is not correct, as there were gigantic protests, civil unrest, and basically mutiny.
But still I think my point sorta stands maybe? I mean:
· Nicholas abdicates,
· he's supposed to be succeeded by his son, a haemophilic child, but Nicholas abdicates on his behalf as well;
· the next supposed successor Michael Alexandrovich gives a weird neither-consent-nor-refusal (I will only accept if this-and-this referendum takes place);
· then everyone sort of decides that this was his way of declining the power;
· and then we're done, and Duma folks like Kerensky take it from there.
It wasn't *quite* the peasants-with-pitchforks situation, in the sense that it was ultimately other political actors who hastened the downfall of monarchy.
How popular was this "popular" revolution?
Wasn't it more of St Petersburg thing?
One contributing factor toward the success of the Iranian Revolution of 1979 may have been that the Shah knew he was going to die soon and didn’t really care about his country’s future.
Most Russians are broadly happy with Putin, and consider him a folk hero. 80% might be an exaggeration, but it's not all that much of one (this is judging by "self-initiative"*, which may be prone to mild exaggeration, but is definitely not going to mask "shy anti-Putin" folks).
*Say, who writes nice-nice things about Putin on their ballot, which is "mostly anonymous" -- this is initiative, and not something taken for social credit.
How many people would "write nice-nice things about Putin on their ballot"? It is probably not even 1%. Most Russians are a depoliticised blob which is neither for nor against anything and only wants to be left alone. The ones that consider Putin "a folk hero" are defo below 30%, I think, not "most Russians".
> Most Russians are broadly happy with Putin, and consider him a folk hero.
This may or may not be true, *but how do you know* one way or the other? The fundamental question of rationality!
> (this is judging by "self-initiative"*, which may be prone to mild exaggeration, but is definitely not going to mask "shy anti-Putin" folks).
Not locally valid.
1. Here's an entire webpage with information on people being arrested for stuff they write on ballots: https://meduza.io/cards/zaderzhaniya-rossiyan-za-nadpisi-na-byulletenyah-pryamoe-narushenie-mnogih-zakonov-vklyuchaya-konstitutsiyu-politseyskih-mogut-nakazat-za-raskrytie-tayny-golosovaniya
You may find other news of the sort just by googling "россия арест бюллетень дискредитация".
People casting a vote would be *realistic* to expect it's unsafe to write pacifist / anti-Putin things.
The point I am making is that "additional inscriptions on ballots" are subject to the same selection bias as Putin's approval rating polls; not only is the refusal rate high, but also the refusal *is more frequent* among those replying one way than the other. People were writing in anti-Putin inscriptions on their ballots in 2024 *despite* risking arrest.
2. Suppose we have somehow found stats about how this many ballots contain pro-Putin vs. anti-Putin inscriptions, and this many are blank. How does this generalize to the whole population? If you predict that most Russians are broadly happy, do you also predict that ballots with pro-Putin stuff are the majority — more frequent that ballots with just a check mark?
3. In any case, do you, in fact, have access to such statistics?
4. The "folk hero" interpretation is what the sequences would call a "burdensome detail". You think a majority of Russians not just approve of Putin, but consider him a folk hero *specifically*? What is this based on? Is that some statistical model I'm not familiar with?
***
I'm sorry if I'm being grouchy. I have spent a lifetime seeing Putin's system try to fake my consent to thinks I actively opposed, and I have had my chances to say fuck you to it, and in a way I am not surprised anymore. What is still beyond me is the phenomenon of buying this whole "he represents them all, that's what they are all like" story that Putin is specifically planting, and this being perceived as brave anti-Putinism.
I don't *actually* know what the distribution of opinions on Putin is like among Russians. But rationality is all we have in an uncertain world; if you care about what society's attitutes are like, it's also necessary to care if your inferences are valid and actually model the thing you think they model.
As someone said, "First they took our epistemology; we don't know what happened after that."
***
I worry I have taken this into culture-war stuff, and I apologize to Scott for endangering his comment section like that. Feel free to remove this comment if you feel it risks bringing in low-effort and/or hateful comments.
Well, to start out with, you have to start with discussing Putin as a Young WEF World Leader... who essentially backstabbed the "sclerotic Western European powers." I can give you more information about this, but... let it be said that there was a time in America when Sarah Palin was laughed at for saying "Russia is The Big Bad" and not very much later, Trump was being accused of "being run by Russia."
This was not an accident, nor was it happenstance, nor was Russia suddenly "a very convenient enemy." There were some very real enemies Putin made, and they were allowed to leave Russia (just, without the money they'd made).
Many Russians give Putin credit for stopping the West (see enemies made above) from thieving (this is at the heart of his reputation as a Folk Hero, you can think of this varietal of Folk Hero as like George Washington). This is above, and beside, Russian sentiment on any particular issue (importantly, right now, on the Ukraine).
<I'm agreeing with you -- my stance is merely that you can probably count "enhanced pro-Putin" ballots as a reasonable floor on "how many people Really Like Vlad">
I do not say that you agree with Putin (in all probability, your posting here increases the likelihood that you'd disagree with Putin...). And, indeed, what I've said above might be more accurate to characterize Putin as "czarlike" (aka personal popularity above that of Russia's government)
I listen to non-Russian sources whose job it is to know this... In so far as anyone can know the minds of over a million people, the data-gathering analysts can figure out who will win the next American Election (once people start deciding, anyway), and they can also figure out how well liked Putin is.
> I listen to non-Russian sources whose job it is to know this... In so far as anyone can know the minds of over a million people, the data-gathering analysts can figure out who will win the next American Election (once people start deciding, anyway), and they can also figure out how well liked Putin is.
This is a helpful clarification. Could you post a link to said sources? I'd like to evaluate them for myself as well.
> the data-gathering analysts can figure out who will win the next American Election (once people start deciding, anyway), and they can also figure out how well liked Putin is.
What I've been trying to convey is that, in my view, there is a deep sense in which predicting who will win the next election in the US is an easier task than figuring out how well liked Putin is inside Russia. At the very least it's a separate category of problems. I'd like to read more first, though.
A personal interview, on this subject at least (also where I sourced the kiwifarms anecdote above, and my source for most Russian/Ukrainian shibola).
I agree, in so far as you can find ways to evaluate "shy Trump" voters, but "shy anti-Putin voters" are likely staying shy for "very valid reasons" not just social opprobium. So it's hard to tell whether folks are "pallid Putin supporters" or "pallid anti-Putin voices" -- as the later have a very real incentive to behave like the former. Vehement anti-Putin voices are probably planning a revolution or something like that (and are probably tracked by intelligence foreign and domestic), so they're actually easy to find, as they have money-paths to foreign entities.
There's another point to be made: in a repressive regime, it is normal for people to "do what everyone else does" (so pallid supporters and opponents in some sense "are" the same -- and propaganda can actualize them into flip-flopping).
> A personal interview, on this subject at least
Okay. I won't update on this, though.
> where I sourced the kiwifarms anecdote above
Sorry, I must have missed it?..
***
A counterargument I found for my own position while looking this up in this discussion is this discussion by Levada Center (one of the places where the 80% support figures come from): https://www.levada.ru/2025/02/26/k-voprosu-ob-oprosah/ arguing that refusal rates for their polling are not significant and are not significantly skewing the outcomes of their polls. I don't know enough to decide if I should buy this or not.
As a mostly-uninformed Westerner (The Berlin Wall fell before I was born but we hadn't scrubbed Cold War concerns from schools yet), the one salient thing to escape was a few years back now, where Russians were being arrested for holding up blank signs. I thought this was brilliant and easily exploitable for image macros, but I couldn't get a sense of how prevalent it was then, before, or since. Since has been the worst, with internet friends in Russia understandably not wanting to get deep into politics and broad disinterest from the media in "the two-week operation has now lasted one hundred weeks"
>What is still beyond me is the phenomenon of buying this whole "he represents them all, that's what they are all like" story that Putin is specifically planting, and this being perceived as brave anti-Putinism.
Let's also make regulations denying them visas and opening bank accounts, that will surely be helpful in defeating Putin
Most Americans are broadly happy with Trump.
Ya sure?
https://www.rasmussenreports.com/public_content/politics/trump_administration_second_term/prez_track_feb06
Rasmussen is a reliable, respected red pollster, that's not "trying to fudge the numbers blue."
46% strongly disapprove, which means 54% are broadly happy with Trump.
"Strongly disapprove" is also disjoint with "somewhat disapprove", which is not what I would call broad happiness.
"Somewhat disapprove" is not shown on either chart in the link provided, but the top chart does show "total approve", which has been topping out at 47% since last November, and sort of hanging around 45% on average, and is probably the closest thing to "broadly happy" for our purposes.
And again, that's Rasmussen, which has a reputation for showing more Trump-favorable results than other polls.
So, it's close-ish to half, but on the less-than side of it.
>that there are resources for which countries compete, and its peoples are an annoyance rather than the what countries are made of
Sure, and one of those resources is "glory", or "status" in rationalist-speak. I'm pretty sure that Putin's secondary motivation is to be a Great Russian Czar. He models himself on Stalin in pretty much the same way that Stalin looked up to Ivan the Terrible.
58: Regarding the startup idea, this falls into the classic trap of “assume that people are honest on dating apps”. Is there any reason why the optimal strategy would be to rate each option one’s honest opinion? I think the overwhelming majority of men would select either 0 or 9 on every swipe. I’m less sure about the optimal woman strategy, but my first guess is that they would use only 0 or 1 (why waste time on someone who doesn’t rate you a 9?)
This is true if a man wants to go on a date with a woman who isn't very interested in him. I can think of several women who I would be interested in dating if they were super into me, but not if they were only sort of grudgingly giving me a chance.
This is true in the abstract sense. In the real dating market that actually exists, there is no pool of single women super interested in the median single man. The options are go on a marginal date and hope to hit it off, or go on no dates. The binary “9 or 0” option dominates except at the very top of the distribution where men are just using the app for large numbers of quick hookups.
Yep, strongly agree, although I think the causation primarily goes the other way. Most women will rate most men very low (0-3) as they don't want to go out with guys who don't really care about them. Then men realise they don't get any dates unless they rate almost every woman highly, and decide to play the numbers game on any moderately good-looking woman. Then the ranking system ceases to convey any meaningful information and you end up with all the same problems as normal dating apps.
I'm gonna be honest - I don't really understand the reasoning process here. Can you elaborate?
I would intuitively think that really mismatched levels of interest would be bad - I don't want to date somebody who is really into me but I feel lukewarm about, because ultimately I'm probably going to disappoint them. And I definitely don't want to date anybody who would only accept because I'm so obsessed with them.
I definitely wouldn't want a big long-term mismatch, but I can see Scott's system making sense under certain conditions. It can be worthwhile to go on a date with someone you don't yet have strong feelings for, as long as you think there's a realistic chance that you might develop them. And if there are a bunch of people in that category, then rather than trying to date all of them, I might want to filter out the ones who are lukewarm about me and give it a chance with the few who are actively interested in me. (If nothing else, the odds of "I develop feelings and she retains them" are probably significantly higher than "we both develop feelings".)
This seems more like an argument for going on exploratory dates. Which makes sense to me! It's kind of silly we try to come to conclusions about how interested we are in people we've never met or barely interacted with. Although nobody seems like they're very interested in a process where you go on 3-5 dates with somebody and THEN decide if you're interested.
Yeah this partly depends on whether we're talking about an app for meeting strangers or an app for facilitating connections with people you already know. Scott said "I can think of several women who I would be interested in dating if they were super into me", so he might be thinking mainly of the latter case, but I'm not sure.
I think if we are talking about people who already know each other, it makes sense that there would be friction here -- like it or not, asking someone out does send a signal of interest, so you have to worry about embarrassment/harming friendships by creating awkwardness/wrongly giving someone the impression that you're into them when actually you're only looking for an exploratory date. And I can see the app helping with some of those things.
For strangers, I reckon you're probably right and it makes more sense just to lower your early-stage standards a bit. (Though I do think it's fine to settle on a 'no' after one meeting rather than feeling obliged to give the person a proper chance.)
Couldn't you just code the app to punish people who do that? A simple algorithm could probably tell if people were trying to game the system
You can’t ungamify dating. That’s like trying to ungamify poker.
I do think there is some possible algorithm that would incentivize an honest ordered ranking of potential partners. Enjoy your Nobel prize when you figure it out.
Econ Nobelist Ken Arrow discussed the problems of honestly ranked ballots in voting... but someone above suggested https://www.astralcodexten.com/p/links-for-february-2026/comment/210331449 just applying a curve to ratings so people who mostly give extreme ratings will have those shifted toward the median.
So the strategy here would be to give the median ranking to everyone except for someone you both like a lot and think you have a chance with. Sounds pretty good, expect that’s already the meta on like-throttled apps.
If the app is matching people with the top X% of their ratings, then a median rating could be lumped in like an extreme one. It does occur to me that people might be incentivized to avoid rating people at all if that could mess with their curve.
Just let Claude do it
If Claude is currently capable of solving this then I need to increase my p(doom) by a few dozen percent.
Surely the big problem is that a ranked ordering of people I've never met is pretty useless? Interest develops over time, after interacting with someone in person.
The Tinder algorithm actually seems optimal here, just a binary no-or-maybe.
Is there any analog to a Vickrey auction (winning bid pays second-highest bid price) that could be used in a dating context? ( Game theory is _not_ my strong point. )
> You can’t ungamify dating. That’s like trying to ungamify poker.
This is one of those things where I see it and am instantly like, "Huh, what is YOUR model for what we're talking about?"
Isn't it as simple as 'You can only see X profiles per day, and you get 5x points to allocate amongst them'?
I suspect there’s a way to do it where, instead of inputting a number, you do something like click higher or lower on the screen, and the app takes a bunch of your clicks and rescales them to give something like a uniform distribution across the numbers. Whatever gaming you might try, the app will undo.
I think this is pretty easy to fix from a design perspective: scale each user's responses so that their personal average is a 5, or give them a limited number of points to allocate across a set of 10 people each week, or etc.
#9, #48 It’s a talent to write as cogently as Cory Doctorow while being so cocksure.
Seems pretty obvious you’d need a lot of subsidy / VC capital yourself to undercut Lyft and Uber as a third wheel with no mindshare and no new gimmick (like autonomous cars)
#47: Naval Gazing has somewhat recently made a post about hydrofoils:
https://www.navalgazing.net/Exotic-Hulls-Part-3
It raises a few more problems, independent of what that company claims to have solved, such as wing erosion through cavitation and floating debris.
Also the side claim of revolutionizing military navies is also most likely just fanciful marketing speak - if anyone can disregard the economies of a technology, it's large militaries, and the large militaries that seriously tried hydrofoils found at most niche uses; improving the technology/economics won't change that. It's like making a sports car that's faster than a Ferrari and costs less than a bicycle, and the military still wouldn't have much use for it.
#55 "Aren’t many murders in very fast attacks where the murderer doesn’t get to choose how many shots/stabs to land?) and would welcome more research on this topic."
Probably important to distinguish between murder (planned) and manslaughter (more spontaneous, "In the moment"). No reason why a murderer shouldn't have enough time to contemplate such things, and in official/scientific crime statistics, I wouldn't expect those terms to be conflated.
To sort of defend theories that this might be militarily useful, we are talking about something a lot smaller and cheaper than previous submerged-foil hydrofoils, which could open up some new roles. They're talking about a 30 kt boat in Sea State 3, which is better than you can do with a conventional 30' boat. Put a diesel in it, and maybe there's some utility as an unmanned surveillance platform, particularly because it can now make normal warship speed. The problem is that I don't think Sea State 3 is enough. "WaterUber isn't running today because it's windy and the waves are too rough" is acceptable as an occasional thing, but it's probably not going to be enough for the military, particularly because it's going to happen a lot more wherever they happen to be. (It's 5 and over 40% of the time in the North Atlantic, which is the only number I happen to have to hand.)
We just need another Russian invasion of Iran to make the Caspian Sea militarily relevant; then these things (and the Caspian Sea Monsters) will finally have their chance to shine.
> 32: 60 Minutes recorded a segment on CECOT (El Salvador torture prison being used by Trump administration), then tried to suppress it (probably under indirect pressure from the administration), then changed its mind and showed it after all.
That is one way to frame the story. Here is a more neutral framing: CBS news hired a new leader (Bari Weiss) to oversee a news organization that lost trust by being rabidly anti-conservative. She noticed that a story about CECOT repeated claims that were at least debatable, at worst false. The reporters involved did not bother to try and understand the administration's position (for example, they made no effort to ask the administration why some choices were made). She told the reporters to at least ask the administration for comment, and refused to air their story until they did so. The reporters assumed that anyone asking for both sides of a story involving Trump must be a secretly working for Trump, and threw a fit.
> I have a bias towards Streisand Effect-ing things that get suppressed like this, so I’ll link it here even though it got on 60 Minutes eventually anyway.
Bari Weiss was a reporter at the NYT. She left (or was kicked out) because she saw nothing wrong with the NYT publishing an editorial by a republican senator (Tom Cotton) arguing that the BLM "mostly peaceful protests" were bad. She stood up for the idea that a newspaper ought to publish things, even if the reporters don't want their audience to hear those things. This was an unforgivable offense at the Times, and she was removed for it.
Her exit is solid evidence that she agrees with you that people should be allowed to speak, and the people who remain at the NYT believe the opposite. Scott, you are repeating a version of this story that cherry-picks facts so as to make a person who is advocating for fairness and reason into the villan.
Backstory on Bari Weiss: https://jessesingal.substack.com/p/congratulations-on-getting-bari-weiss?utm_source=post-email-title&publication_id=4833&post_id=175455359&utm_campaign=email-post-title&isFreemail=true&r=5hi9n&triedRedirect=true
I don't think "person believed Republicans should be allowed to speak in 2020" is solid evidence for "person is willing to criticize the Trump administration in 2025", and I think there are many, many counterexamples. I think she's amply displayed an anti-free-speech streak since then especially in her dealings with Israel critics.
Given that CBS has been trying to suck up to Trump ever since he gave them permission to do a controversial merger in exchange for (reading between the lines) being more pro-administration, I don't think it's a crazy theory to believe that the person hired to be pro-administration by their new pro-administration leader would quash an anti-administration story.
Can you present evidence for your claim that she told the reporters to ask for comment and they refused? What I see on Wikipedia is that "Alfonsi said Weiss had not contacted her before spiking the story and journalist Scott Pelley charged that Weiss had not attended any of five internal screenings of the story during the final stages of editing."
I said in the post that it was "indirectly suppressed", by which I meant Trump was tweeting that he didn't like CBS, everyone knew that the CBS merger depended on his good feelings, and so they complied in advance with what they knew his desires to be. I agree he didn't specifically send an email saying "cancel this".
They've also shown things that don't get the government's perspective, so I don't think "you can think of one 60 Minutes segment that doesn't fit this story" necessarily proves anything.
It wasn't "suppressed", it was delayed.
Only a full suppression or a complete redo would have benefitted the administration. Hard to see what it got out of the delay, other than an opportunity to comment which they apparently did not take.
Right. And the idea that Weiss only released it because she was "forced" (even though CBS announced that it would be aired later) implies there's someone both capable and willing to force her on that, like if David Ellison had a road to Damascus moment.
I don't think it's a completely crazy theory a priori, but CBS's overall behavior recently (including how they've covered other stories) looks a lot more like "trying to be more honest and rein in biased reporting" than "bowing to arbitrary admin pressure to bury stories".
Scott Said:
"""
Can you present evidence for your claim that she told the reporters to ask for comment and they refused?
"""
This article has the full text of the memo she sent about the story to people inside CBS (scroll to the end to see it). In that memo she points out several ways the report doesn't meet basic standards of fairness:
https://nypost.com/2025/12/22/media/read-the-memo-bari-weiss-sent-to-cbs-staffers-after-yanking-60-minutes-spot/
As evidence that she told reporters to ask for comment: Bullet point two from that memo:
"""
At present, we do not present the administration’s argument for why it sent 252 Venezuelans to CECOT. What we have is Karoline Leavitt’s soundbite claiming they are evildoers in America (rapists, murderers, etc.). But isn’t there much more to ask in light of the torture that we are revealing? Tom Homan and Stephen Miller don’t tend to be shy. I realize we’ve emailed the DHS spox, but we need to push much harder to get these principals on the record.
"""
As evidence they refused: The video was leaked. If the people involved took the feedback in that memo seriously, and wanted to address it, they would do so before showing the video to the world.
It is possible that the some video editor or IT person chose to leak the video without the reporter's knowledge, and against their wishes. Given the way this is being reported (search google news for "Bari Weiss CBS 60 minutes"), it is clear that the more common view among reporters is the one you shared: Weiss is biased in favor of Trump, and asking reporters to be even-handed is evidence of her bias.
Scott Said:
"""
Given that CBS has been trying to suck up to Trump ever since he gave them permission to do a controversial merger in exchange for (reading between the lines) being more pro-administration, I don't think it's a crazy theory to believe that the person hired to be pro-administration by their new pro-administration leader would quash an anti-administration story.
"""
Why do you believe she was "hired to be pro-administration by their new pro-administration leader"? It seems more likely to me that she was hired because the existing reporters let their politics color their reporting.
I can't read her mind. I can read her memo, and it seems to me to be a well-reasoned request to make the reporting less biased. Do you disagree? If so, what part is unfair?
The main reasons I don't trust that memo as being the full story (all taken from https://en.wikipedia.org/wiki/Inside_CECOT):
> "On the afternoon of December 21, 2025, [three hours] prior to the airing of the episode, CBS News announced the CECOT segment had been pulled from the episode, and would air at a later date, without disclosing a reason. Multiple sources and commentators noted that this action was almost unprecedented, with stories almost never being pulled after such a vigorous screening process as typically occurs at 60 Minutes"
> "Alfonsi stated in an email memorandum to colleagues that the production team did request comments and interviews from White House officials, the State Department, and the Department of Homeland Security (DHS, which were all refused). And that the segment was reviewed five times and approved by the CBS legal department and Standards and Practices division."
> "Alfonsi had requested an interview with Secretary of Homeland Security Kristi Noem roughly a month before the feature was to be broadcast, but received no response; following additional requests for comment made to their office, DHS referred all questions to the Salvadoran government. Axios has reported that the State Department provided comment as well. In a response sent on the 18th, the Trump administration questioned why 60 Minutes had not focused their efforts on the experiences of Angel Parents, "whose innocent American children [had] tragically been murdered by vicious illegal aliens that President Trump are [sic.] removing from the country."
> "'Government silence is a statement, not a veto. Their refusal to be interviewed is a tactical maneuver designed to kill the story,' she said."
> "At the meeting, 60 Minutes correspondent Scott Pelley again brought up the fact that Weiss had not turned up to any of the five internal screenings of the segment during the final stages of editing. Pundits have noted that her intervention came only after the piece had been finalized and promoted on social media."
So the impression I get is that the people writing the story tried to get both sides and comments from government officials, they presented the story five times in a way that Weiss normally would have had an opportunity to see and comment on, they got it reviewed by the procedure for reviewing these things to make sure it was fair, at each step everyone agreed that since the government officials had refused to comment, it was fair to air it without their comments, since they didn't want to make the government able to trivially crush any story by refusing to comment on it. Then three hours before it was broadcast, Weiss, who should have been monitoring this entire process, suddenly cancelled it in a way that had basically never happened before, saying it needed to get a government official's comment. So either Weiss is extremely incompetent or she was acting on some sort of political orders.
But I've edited your comment and the resulting thread into the main post so people can read it and get a more thorough view of the situation.
I’m not sure how reasonable it would be for the head of CBS News to be present in preliminary internal screenings of 60 Minutes segments. Definitely weird to pull such a hard-hitting segment on short notice, but if Weiss thinks that the internal review departments are incompetent, then her overruling their decision to green-light the segment isn’t necessarily evidence that she was pressured, just evidence that she has different standards than them.
I am preregistering my bias as generally anti CECOT and Trump skeptic, fwiw.
Nevertheless --
I believe the most plausible story accounting for somewhat charitable readings of all sides is that she did not like the story, but that her memo was still some rough approximation of the truth, and she wasn't conspiring to be responsive to trump's whims while lying about it.
I have high uncertainty whether she didn't like the story because, as she said, she did not think it broke new ground relative to other existing reporting, or because she has an arbitrarily high bar for all stories against the administration, or because of some middle ground where she has an arbitrarily high bar for stories that pile on issues already litigated by other outlets. (I find the last of these most likely because in my experience people instinctively soften bad news a *moderate* amount, but not typically to the point of outright lies, so both extremes of perfect honesty and maximal deception seem less probable to me, and this was the only middle ground I was clever enough to imagine.)
She clearly has communication problems with her staff, though no wonder, fostered by the perception that she's an outsider brought in to "correct" their political views, which are no doubt sincerely held, or even personally interpreted as objectivity amidst a hostile political environment.
I downrate the idea of a premeditated spiking because that is inconsistent with her lack of earlier involvement when it would have been easier to inject delays, inconsistent with her lack of apparent control over leaks of the story or leaks about the story's spiking, and inconsistent with her claims in the memo that she would help them get the story over the bar by providing personal contact info of other officials.
Offering to help is a form of expensive signalling. If someone takes you up on it and you do not offer any meaningful assistance it raises the likelihood they will move you from "trembling hand" to "spiteful defector" in their assessment of your behavior, particularly if iterated. I understand that it's possible her game theory is stalling for time, that is plausible but seems less likely to me than her behaving like a normal human in a contested environment.
I do not have good baseline information about how likely it is for a person in her position to watch a story five times in preliminaries. Maybe she was seething with rage at the very idea of watching the story she hoped someone else would spike so she wouldn't have to. Maybe her position doesn't usually do all those reviews. Maybe she's new and should have been there but doesn't actually realize that's a line level expectation (possibly exacerbated by the friction of her presence which are preventing open upward feedback).
One of the most interesting pieces of evidence in the aftermath is that nobody appears to have been fired over this. That raises my p() that they are an organizational unit with many strong opinions but genuinely trying to muddle through from both sides, and lowers my confidence in the idea that they have crossed some kind of us or them rubicon internally. (Though... I'm not sure this is fair. I still expect someone to be let go / pushed out in the next six months after enough time has passed it seems unrelated to this and not sure from which side. Would that make me more convinced this was actually a gutter fight where everybody was playing dirty? Maybe it should proportionally, hmm.)
The facts you are presenting are an account of what happened by people who are not disinterested parties. We have no way to figure out if these claims are true, or are the whole truth. It comes down to the credibility of the people involved. That will never give us proof, but it is the best we can do with limited information.
Scott, as I read your response I think of your essay on "Trapped Priors": You seem to start with the assumption that Weiss is biased in favor of Trump, was installed because she is biased in favor of Trump, etc. I don't share that assumption. I think she is an old-fashioned liberal (to the left of 75% of the American public). I think the reporters are the new sort of democrat, the kind that does not tolerate dissent from the party line (to the left of 95% of the American public). I can't know for sure which of us has the wrong prior.
To me, the following factors weigh in favor of Weiss:
1. On at least one occasion, Weiss stood up for the principle that a news organization should allow editorials that the reporters do not agree with. She did this even when it cost her a job. Very few people can say this. That tells me she has some integrity. It is possible that she was a secret republican mole who went to work for the NYT so she could get canceled to fake credibility, but I doubt it.
2. The memo Weiss wrote is well reasoned and seems to be a genuine attempt at making the reporting fair and honest.
3. The reaction to that memo by journalists generally makes no effort to consider anything she actually said. If the reaction was along the lines of "I can see her point about ..., but I disagree because ..." then I would be inclined to think the reaction is a serious attempt to get a good result. I don't see that. At all.
4. I watched the leaked piece. The issues Weiss raises are, in my opinion, real issues that should have been addressed.
5. I have worked in orgs that were failing, and had a new manager put at the helm. The people from the failing organization hated the new manager, and did everything possible to undermine her. I bet the long-term CBS reporters HATE the fact that someone kicked out of the NYT for not being far-left enough made her own news organization, beat them at the job of producing reputable news, and now has the gaul to tell them what to do.
Isn't it undisputed that the journalists tried to get the administration to offer their side of the story, and the administration refused to do so? That would seem to entirely negate the force of the objection that it would be irresponsible to run the story without getting the administration's side.
In the quote above, Weiss says there are people in the admin who aren't shy and she'd expect to comment, so CBS should push harder.
My impression from this is that the previous attempts had been checkboxing "we (hostile media org" CCd an admin official knowing they don't like to talk to us", and Weiss said "no, I want you toake a real serious effort to get their side of the story here".
1. I still don't think this proves anything. The Communists protested when people suppressed Communism, but that doesn't mean they were generally free-speech loving people who practiced free speech when they were in power.
2. I agree the memo is the best evidence for Weiss' side, but good writers can write memos approximately that well-reasoned and genuine-sounding for basically any scandal you give them. It's not hard for a writer of Weiss' talent to write a memo that says "I was right and doing reasonable things".
3. I think the journalists' response to the memo is "She says she cancelled because we didn't get an administration interview, but we tried to get an administration interview and they refused, and there are known procedures for dealing with this and she ignored all of them in an unprecedented way". I think this is a strong response.
4. Fair enough, I didn't watch the piece.
5. I think this requires that we throw out the testimony of everyone except Bari because they might be biased, while also trusting that Bari is the single unbiased individual in this entire affair (even though of course she comes into this fraught situation with biases of her own). I agree there's a lot of uncertainty because everyone has reason to spin things in their own direction, but I don't see a reason to trust Bari more than anyone else.
But fine, you've convinced me there's some possibility it wasn't politically motivated, though I still think <50%.
Is Bari Weiss analogous to those communists? My understanding is that she has called herself “center left on most things” and “left-leaning centrist”. The Free Press under her leadership has been willing to criticize the Trump admin: https://www.thefp.com/p/is-donald-trump-breaking-the-law
Weiss’ previous integrity is definitely impressive. But between them and now, she worked out an agreement where CBS bought her company. I don’t know much about the integrity of most people in that position, but the most salient example I can think of is from Australian 2010s politics. The opposition leader Malcolm Turnbull resigned when he realised he couldn’t lead the (right-wing) Liberal party to act on climate change. But later, he came back into office by agreeing to give up and accept the party’s anti-climate-action position.
Possibly this is just the media I see. Do you know of stories where people made these principled stands, left, then came back to an organisation which still rejected those principles, and managed to implement the principles anyway? I’d like to see those stories, for their own sake.
> On at least one occasion, Weiss stood up for the principle that a news organization should allow editorials that the reporters do not agree with. She did this even when it cost her a job.
Bari Weiss resigned from the NYT. She wasn't fired or kicked out for standing up for the Tom Cotton op-ed.
Has she ever stood up for the principle of free speech in favor of a position more liberal than her own view?
If it was to be aired at a later date then it wasn't "suppressed", it was delayed.
I'm banning you for saying false things in any overly confident and hostile tone. For my response to what you're saying, see my discussion with Frange above.
I think a lot of the issue people are taking here is that the contentious framing (which I maintain is inaccurate, as Frange stated) comes across as selective bravery. A lot of influencers in this sphere are happy to 'stand up to' people so long as the people they are standing up to aren't the people who tend to retaliate harshly when people do so, and that creates all the wrong incentives.
>I have a bias towards Streisand Effect-ing things that get suppressed like this, so I’ll link it here even though it got on 60 Minutes eventually anyway.
I do not think you would use a similar tone when discussing, for instance, the Rotherham coverups. In fact, when you did mention it, back in 2014, you did everything possible to be 'fair' to both sides of the issue! (https://slatestarcodex.com/2014/10/16/five-case-studies-on-politicization/)
This doesn’t sound like a neutral framing. CBS was only losing trust in the general sense that all legacy media organizations have been losing trust as people move to distributed media. There was no sentiment among anyone that CBS was “rabidly anti-conservative”.
Bari Weiss had a very prominent career in between the New York Times and CBS, and understanding what she was doing in that career is relevant to understanding why she might have gotten hired at CBS.
>There was no sentiment among anyone that CBS was “rabidly anti-conservative”.
I'm pretty sure that plenty of non-leftists think that all of the purportedly neutral media outlets are rabidly anti-conservative.
I would think that it’s only rightists who think that, not most non-leftists. But even they don’t think CBS was worse than ABC or NBC (let alone print media).
Trump singled out CBS in particular as being biased against him during the campaign, after 60 minutes ran that interview with Kamala Harris that was edited to make her sound better. Trump went off on how biased and bad CBS is many times, publicly, in campaign speeches to his voters. He sued, and CBS ended up settling. So it wouldn't be unreasonable for someone, even if they're a centrist swing voter, to get the impression that CBS has an anti-Trump bias of some kind. It was a narrative that was going around.
Ah interesting! I hadn’t noticed that he was discriminately attacking a single media organization rather than just indiscriminately attacking lots of them!
Well he attacked other ones too, but from the date of that 60 minutes interview until Election Day he made CBS his main target in the media.
I gotta say that the "neutral framing" here is very favorable to Weiss, and not entirely accurate.
Cotton's op-ed was not that the protests were bad, it was that the military should be deployed domestically to clamp down on them. It was literally (and accurately) entitled "Send In The Troops".
https://www.nytimes.com/2020/06/03/opinion/tom-cotton-protests-military.html
(Also I honestly am not sure why you feel like this is the main reason she left the NYT, rather than one of many she herself cited. https://www.bariweiss.com/resignation-letter)
But I'll give another reason why the NYT shouldn't have published it: because it's not 1995. Tom Cotton can publish anything he wants at no cost and be confident it will be read by a decent number of people. The NYT doesn't need to signal-boost his opinions if they don't think they have merit, because the days when something needs to be in the NYT to be read are long dead. They should have an ounce of selectivity about which things they print.
Also putting scare quotes around "mostly peaceful protests" is kind of pointless. The protests after George Floyd died were MASSIVE, with tens of millions of participants. They can both be mostly peaceful and, in aggregate, cause a lot of damage.
Also, framing Bari Weiss as a free speech warrior of some kind is just not credible. She's been hounding people for wrongthink on Israel since she was in undergrad. Weiss (like mos people, honestly) talks about the importance of free speech for ideas she supports or agrees with, and turns around and suggests that ideas she doesn't like are unworthy of consideration.
For those interested, the New Yorker recently did a profile of Bari Weiss, including a discussion of the 60 minutes controversy. I like the New Yorker, and I believe their stories to be well-fact-checked, but I understand if others feel they are biased. Regarding this story, they don’t add much beyond what Scott said here, although it does mention that she had offered some earlier suggestions to the story that were incorporated, before later cancelling/deferring it, which weakly supports the idea that the cancellation could have had outside/political motivation: https://archive.ph/4gWia
“””
A few days before Christmas, “60 Minutes” was set to air a report on cecot, the prison in El Salvador where the Trump Administration had sent more than two hundred deportees in March. The men, most of whom were from Venezuela, had been spirited out of the U.S. on a series of late-night flights, in violation of a federal judge’s court order. Sharyn Alfonsi, a “60 Minutes” correspondent, had interviewed two of them. The stories were harrowing. “There was blood everywhere, screams, people crying, people who couldn’t take it and were urinating and vomiting on themselves,” one of the men told her. She had reached out to the Trump Administration for comment but received only a cursory, two-sentence response.
Like all “60 Minutes” stories, the cecot segment had been fact-checked and vetted by the network’s legal department. Five separate screenings were held for various editorial stakeholders. Weiss was supposed to attend the final screening, on Thursday afternoon, but she had missed it. She didn’t see the segment until late that night, e-mailing suggestions for a few changes that were incorporated into the piece. The network promoted the segment as the lead story for that Sunday’s episode. Alfonsi flew home to Texas.
The cecot story was being finalized almost exactly as Ellison was pursuing another major expansion of his growing media empire. Earlier that month, he had launched a hostile takeover bid to purchase Warner Bros. Discovery, the film-and-television conglomerate that had already announced a deal to be acquired by Netflix. In many ways, the outcome depended on Trump, since regulatory approval would be required for either sale to go through. “I’ll be involved in that decision,” Trump had told reporters.
For Ellison, this was suddenly a problem. “60 Minutes” had recently aired an interview with Marjorie Taylor Greene, a former Trump ally who had fallen out with the President over his resistance to releasing the F.B.I.’s files on the convicted sex offender Jeffrey Epstein. “they are no better than the old ownership,” Trump posted on Truth Social after the interview aired. “Since they bought it, 60 minutes has actually gotten worse.”
On Saturday morning, Weiss, who reports directly to Ellison, told “60 Minutes” producers she was concerned that no officials from the Trump Administration had been interviewed on camera for the piece. Specifically, she wanted to include the Administration’s argument for its use of the Alien Enemies Act, an eighteenth-century law that Trump officials claimed allowed them to deport Venezuelan migrants to El Salvador without due process. “We should explain this, with a voice arguing that Trump is exceeding his authority under the relevant statute, and another arguing that he’s operating within the bounds of his authority,” Weiss wrote in a note to the producers. “There’s a genuine debate here.” She had “tracked down” numbers for Tom Homan, the border czar, and Stephen Miller, Trump’s chief immigration adviser, which she sent to the “60 Minutes” team.
On Sunday, three hours before broadcast, the cecot piece was officially pulled from the lineup. “My job is to make sure that all stories we publish are the best they can be,” Weiss said in a statement. “Holding stories that aren’t ready for whatever reason—that they lack sufficient context, say, or that they are missing critical voices—happens every day in every newsroom.” Alfonsi felt betrayed. “In my view, pulling it now, after every rigorous internal check has been met, is not an editorial decision, it is a political one,” she wrote in a note to her colleagues. “Government silence is a statement, not a veto. Their refusal to be interviewed is a tactical maneuver designed to kill the story.”
The next day, the Ellisons announced a significant sweetening of their bid for Warner Bros.: Larry would personally guarantee $40.4 billion of the funding. Weiss didn’t come to the CBS offices that day. She joined the morning’s daily editorial meeting via Zoom, beginning with what seemed like a rebuke of Alfonsi. “The only newsroom that I’m interested in running is one where we are able to have contentious disagreements about the thorniest editorial matters and do so with respect, and, crucially, where we assume the best intent of our colleagues,” Weiss said. “And anything else is absolutely unacceptable to me and should be unacceptable to you.”
That afternoon, during a “60 Minutes” staff meeting, Scott Pelley, a longtime correspondent, expressed frustration that Weiss hadn’t attended any of the screenings of the segment or communicated directly with Alfonsi. “She needs to take her job a little bit more seriously,” he said. A former CBS staffer was soon circulating an open letter to Ellison, expressing alarm at “a breakdown in editorial oversight” that risked “setting a dangerous precedent in a country that has traditionally valued press freedom.” A former CBS executive told me that, even if Weiss’s concerns had been valid, her decision to cut the segment at such a late hour had opened her up to charges of corporate interference: “It makes you wonder, Did someone call once they saw the promo on the air and then she spent more time on it because there was some big complaint?” Sources close to Weiss and Ellison said that Skydance leadership had zero involvement in the story and did not screen the piece.
These sources also told me that Weiss “readily realizes and admits that she was not as knowledgeable as she should have been about the timing of the marketing and promo process at ‘60 Minutes.’ She brings the sometimes chaotic energy and work ethic of a startup, but she also realizes she needs to work on having more executive discipline.” Weiss also seemed to be struggling with the fact that, at a time when the Trump Administration is routinely lying to the public and straining to justify blatant abuses of executive power, often with violent or deadly consequences, she was still wedded to the idea of news coverage as a contest of ideas, in which both sides of the debate are equally valid. Privately, she has expressed alarm at many of the Administration’s actions, a person close to Weiss told me. But, in her role as the editor-in-chief of CBS News, her main concern is being able to book its main players on her network’s shows.
“””
#4:
The dynamics of the Manifold COVID origins market are explainable based on three factors:
(A) Market structure that'll likely never resolve, leading to only a small number of motivated individuals dominating the trading; (B) financial changes in Manifold affecting the behaviour of those individuals; (C) user churn among them
(A): As you note, the market structure is such that it is very unlikely to resolve (ever).
This means that people who I'd consider to be "serious" traders don't really participate much in this market on a volume basis - it's just not clearly profitable when your mana will be locked forever and price isn't really tethered to reality by an outcome happening.
The market has had 4.6m in mana volume; about 25% of the total is Peter Miller alone on the zoonosis side and about 15% is the top 3 traders on the lab leak side.
In contrast, the biggest trader by position I'd consider 'serious' I can see in the top lists has about a 16k zoonosis position, or 0.34% of volume.
When I say the top holders aren't "serious", I don't mean this as a diss to them personally, just that they're not necessarily the most profitable users on site, and they don't trade a lot on other unrelated markets.
(B): Manifold keeps on iterating on its financial system, in particular the loan system. Over the last year or two they've removed and recreated the loan system a couple of times, as well as adding per-market leverage caps and reducing loan dispersion rates. This means that there's less liquidity flowing for single-issue traders to stake on this market; and when there are relatively few major participants, this matters a lot.
(C): Some of the biggest holders on the lab leak side have simply lost interest. "Mike P", the biggest holder on the lab leak side, used to trade ~daily and appears to have quit Manifold in early 2025; "Robert Sutherland", the second biggest holder, traded frequently through April and now only does sporadically, I'm guessing substantially due to the loan system changes.
Basically, the whole market is "Mike P, Robert Sutherland, and George Riddick (lab leak) vs Peter Miller (zoonosis): who will blink first?" and so far it seems that Peter Miller has more staying power.
This is extremely helpful and, at a meta level, a useful framework for understanding how much to weight other prediction markets.
I think the dynamics of this market are relatively unique -- most markets on Manifold don't have this particular issue, and it doesn't apply at all on other forecasting or prediction market platforms, where there's a strong editorial voice ensuring markets are in principle resolvable.
The part that is most transferable is looking at forecaster P&L for top traders on a question in a real-money market, although I'd note that that's also not without thorns and shouldn't be taken naively.
E.g. it's very common for sophisticated traders to hedge across correlated markets, so just because you see a profitable user taking a big position on one side of a particular question, does not mean that they necessarily strongly believe that side.
Maybe I should have been praising the meta meta level of your analysis...
Maybe every market needs an auto-generated secondary market on whether the original market will ever resolve...
How was it ever expected to resolve?
I'm not sure if I understand the question you're asking.
If it's literal, "how can this market resolve?", I guess it could in theory if a lab leak smoking gun was discovered, or if a complete zoonotic transmission chain were documented in a high status scientific publication.
If it's emphatic, "why do people trade at all when obviously it won't resolve?", idk, trading is fun and the debate was spirited
If it's technical, "how could one write a prediction market like this that would resolve?", you can ask about a proxy and specify a cutoff time, e.g. will a paper be published in a top tier journal that concludes one way by date XYZ.
I mean how was the resolution criteria for the market written?
The point of Manifold is that creating vague markets is good, and easy, vs. not creating markets.
It could be vague, but I was curious about what was actually written. So I looked it up and here is the section I found relevant:
This market resolves once we have a definitive answer to this question. (i.e. "I've looked at all notable evidence presented by both sides and have upwards of 98% confidence that a certain conclusion is correct, and it doesn't seem likely that any further relevant evidence will be forthcoming any time soon.")
This will likely not occur until many years after Covid is no longer a subject of active political contention, motivations for various actors to distort or hide inconvenient evidence have died down, and a scientific consensus has emerged on the subject. For exactly when it will resolve, see When will the Covid lab leak market resolve?
I will be conferring with the community extensively before resolving this market, to ensure I haven't missed anything and aren't being overconfident in one direction or another. As some additional assurance, see Will my resolution of the COVID-19 lab leak market be controversial?31%
(For comparison, the level of evidence in favor of anthropogenic climate change would be sufficient, despite the existence of a few doubts here and there.)
If we never reach a point where I can safely be that confident either way, it'll remain open indefinitely. (And Manifold lends you your mana back after a few months, so this doesn't negatively impact you.)
Hi, I have the third largest zoonosis position on that market. I agree with you directionally, but also want to push back a bit on your framing.
For example: My position is indeed much smaller than Miller's and I don't think of myself as a serious trader (proof: serious trader should trade on other markets). But I'm not a noob either — e.g. see my profits/calibration https://manifold.markets/VaclavRozhon/calibration . I just care more about testing my calibration than winning mana. I can see more semiserious people with semiserious stakes in this market. So there should be some signal there.
Then again, I of course mostly agree with you and don't really want to claim that this market is somehow efficient because I don't think it is -- if serious people traded, I think lab leak would have crashed much sooner.
I appreciate the pushback. I admit I didn't really rigorously check the whole list for 'seriousness', I was mostly just looking at the usernames to see who I recognised, assuming I'd recognise more or less every 'serious' trader. But you actually have more lifetime profit than I do! So, clearly, this method is imperfect.
FWIW, you're at least half right. There was a bidding war for a while. The market crashed after Scott's post about covid origins, some large users like Joshua started asking why not just take it down to 10% to match Scott's views. Mike P started buying mana to prop it up, I think he spent ~$300 on that until the No bettors gave up.
I had previously steered clear of any large bets in the market, but at some point I decided to just see if I could overwhelm Mike and I got it back down below 50%.
I mostly gave up after that contest -- you can see my average price in the market is 51%. I've thrown a few smaller bets in below 50%, but not so much compared to fighting around the middle. If it was just me, I probably would have just thrown in all my mana as a wall at 49% and called that good enough.
Other people have since decided to take the market well below 50%. I don't know their motives, or what they are expecting as far as resolution. That could either be a shift in sentiment against the lab leak theory or a shift in the average participant in the market.
FWIW, I think the actual odds that Covid is a lab leak are non-zero but less than 1%. Others disagree, and I think the market has little connection to any real assessment of the odds, either way.
I'm not even sure what the rational price should be. Should it match expert surveys and settle somewhere in the 10-20% range?
Even if it were certain to resolve as No in 2040, what should the price be based on the risk free rate? People will bet on the UFO market at 95% for a guaranteed win in ~2 years, so should it be 43% return for 14 years? Or maybe less because it could potentially resolve sooner?
Worth emphasizing: The market dropped after *Scott's post* about the Rootclaim debate, not after the Rootclaim debate itself. This, I think, proved that the market probability was not being determined by informed actors.
What’s the mimetics of weaboo? Terms don’t escape containment unless they have a reason to spread, and weaboo doesn’t seem particularly clever to the average folk without context. Is it because it’s fun to say and vaguely Japanese sounding?
IME most people say "weeb" now. This is shorter and resembles the older "dweeb", which carried similar pejorative connotations.
It's because wapanese/weeaboos/weebs/whatever are a real phenomenon, significant enough to be regularly discussed in certain circles (basically anyplace anime/manga-adjacent, for starters). which means there's strong incentive to find a short memorable name. The first candidate was obvious and descriptive but sounded like it might be sort of racist, so as soon as there was any significant usage of the new "means the same thing as wapanese but Officially Not Racist" term, it took off. I had not known that 4chan was the source of this linguistic evolution, but it makes sense.
And the meme continues to spread. We now also have "Wehraboo" (from "Wehrmacht"), for people who are as gushingly enthusiastic about WWII-era German military culture as the Weebs are about Japanese culture.
We do not have "Weeaboo II: Electric Boogaloo". Yet.
We also have "Teaboo" for someone obsessed with British culture, and "Ouiaboo" for their French equivalent.
And "freeaboo" for someone obsessed with American culture.
And Ouijiboo for those obsessed with the spirit world.
Also Koreaboo!
The first time I heard the term Koreaboo, I asked the person who used it if it meant a white person obsessed with Korea, or a Korean person obsessed with Japan. He told me it meant the former, and added that the word for the latter was "traitor".
What was the first obvious-and-descriptive-but-sort-of-racist term?
"Wapanese", short for "white Japanese". which explicitly mingles two race-labels and adds a derogatory spin.
Before you know it, we'll have so many people gushingly enthusiastic about calling out other embarrassingly enthusiastic fans that we coin "Electric Boogaliboo".
Regarding item #10, I'd like to recommend the book *Wild Geese Returning: Chinese Reversible Poems*, by Michèle Métail (a member of the Oulipo, for those who know what that is), originally published in French in 2011 and translated by Jody Gladding in 2017. It's about the whole genre of poems that Su Hui kicked off, which persisted for centuries. About 1/4 of the book is about Su Hui's poem, including some translated sections (as far as I can tell, the whole poem hasn't been translated; but as far as I can tell the project would be practically and maybe theoretically impossible). And then it includes a whole lot of other reversible poems. A great book.
I don't seem to be able to access the Claude link for #60, is it just me?
Should be fixed now, can someone confirm?
Confirmed!
>The only crime that isn’t like this is murder - everyone notices a missing/dead human, and the police have to investigate all of them.
One could imagine a dynamic where the police itself becomes lazy and uninterested, and would rule potential homicides as accidents or suicides to save themselves from the hassle of investigating it.
Or because their funding/promotions/etc rely on hitting certain metrics.
I basically don't trust any statistic from a police department for the same reason I haven't trusted standardized test scores since No Child Left Behind: the measures became metrics, there's overwhelming incentives to fudge them and little oversight to catch it.
Adjacent example on the BBC today. Police assumed that a dead body they found was non suspicious (they thought he'd fallen). Only later, days later, did they realise he'd been shot with a shotgun.
https://www.bbc.co.uk/news/articles/cp8283qj4gro
There was a similar example I saw on some TV show a while back, where it appeared a man died alone in a hotel room without any injuries, but during the autopsy, they found out his heart exploded. Turns out he got accidentally shot in the crotch from the room next to his, and the bullet entered in such a way that it closed the wound behind itself.
58. Matches a Jordan Harbinger discussion about women giving a hard no to men under 6' on dating sites, when it's actually a weak preference.
38. Yikes. Here’s an excerpt from the description:
“If your ancestors aren’t entitled to control your beliefs and practices, why are you entitled to control the beliefs and practices of the strangers who surround you? The best cultural policy is not cultural preservation, but cultural competition.”
I’ve seen Bryan Caplan’s view on this stuff, and I’m not sure I 100% understand it, because at face value it seems obviously wrong. If a culture of cooperation meets a culture of defection (I.E. Welfare being generally used as-needed vs. welfare being maximized and gamed to produce the most money for the least amount of work) the defecting culture will “outcompete” the cooperating culture.
In liberal democracies, with welfare, it seems impossible for the open-borders world Caplan advocates for to function. Population equates to political power, and there is little incentive to *not* game welfare when you have no preexisting social bonds to the society you’re receiving welfare from.
42. AI generated images of Maduro were being broadcast on CNN and Fox News for days after his capture. https://www.cbsnews.com/amp/news/maduro-capture-ai-images-false-information/
I feel like this was a much broader instance of AI-trickery, since that Uber Eats thing was only interesting to a specific corner of the internet, while probably half of America saw the Maduro pictures on the news and thought they were real. Of course the impact was smaller, since it wasn’t really misrepresenting the situation at all.
1. Is it actually true that modern welfare states are "cultures of cooperation"? Or do they just have strong institutions that make it hard to cheat? ("Welfare queens" have been a stock boogeyman for Republicans for decades, regardless of what the immigration rate is.)
2. Is it actually true that the "social bonds" you get from being native to a society disincentivize antisocial actions? If so, why is the crime rate for immigrants *lower* than for natives?
(Also, while "social ties discourage defection" is an argument that makes sense for not doing crimes to your friends and neighbors, I'm not sure the same argument applies when you're defrauding the distant and impersonal federal government.)
3. If an immigrant plans to live in their new country forever, don't they have the same incentive as any native to ensure that the country they live in stays strong?
1. Is it actually true that modern welfare states are "cultures of cooperation"?
Hard to tell. The most successful states with expansive social welfare have high rates of social trust: https://www.visualcapitalist.com/mapped-where-people-trust-each-other-most-by-country/
2. Is it actually true that the "social bonds" you get from being native to a society disincentivize antisocial actions? If so, why is the crime rate for immigrants *lower* than for natives?
Filtering and a higher baseline crime level would be a strong culprit if you're talking about the US. In Europe migrants are significantly overrepresented in number of crimes.
3. If an immigrant plans to live in their new country forever, don't they have the same incentive as any native to ensure that the country they live in stays strong?
Not if gaming the system benefits you personally. There is a very strong incentive to take free money for you personally, even if in some abstract and minuscule sense contributes to harming the nation.
---
But take this to the extreme as an example. To be eligible for SNAP in the US, a 3-person household needs to make less than $20,352, meaning that ~95% of the global population would qualify for basic US welfare, not even considering benefits that extend well above the $20k maximum (larger households, housing assistance, etc.).
Living on welfare in the US doesn't give you a great quality of life compared to what you can get by working. If you're able to, most people work. Living on welfare in the US compared to much of the world gives you a great quality of life, and you also don't have to work. Open borders, with the sort of welfare systems we currently have seems incompatible to me.
And the cultural argument seems obviously wrong too. If a hundred fundamentalist Muslims move to your town and have 5 children on average, while your population has 1.5, over a few generations your community will become Muslim. Sure, there will probably be some defections over to Western Liberal society, but in the meantime you will get significant conflicts of values that can cause serious problems, and in the long-term your cultural values may be less "competitive", so they lose.
And I'm a very pro-immigration person. I think that for persons with the means to sustain themselves the bar should be basically "Show up with a clean criminal record, and you're given a SSN." But then Caplan comes in with his advocacy for complete open borders and I feel that my (IMO internally consistent) view becomes anti-immigrant. IK Caplan is a libertarian and doesn't support the welfare state in general, but his advocacy for this single issue, without clarifying that it probably wouldn't work unless we overhauled a large part of the rest of society as well, is pretty bad.
It should probably be pointed out that the welfare state isn't the actual issue. The reason for migration in the current age (and, to be fair, in the past as well) is overwhelmingly the Bank Robber's Maxim: that's where the money is.
If we stop handing out money, in the form of welfare, that does not necessarily stop the immigrants from coming - not if they see a different way of effecting a move of money from someone else's pocket into theirs (hint: I am not talking about hard work).
Though coming to America and working hard is in fact an excellent way to move money from someone else's pocket into your own. Often much more effective than staying in your home country and working hard.
Absolutely! However, this assumes that there will be a job for you, somewhere. Currently, that's still a decent bet in the US. In Europe, the situation isn't as rosy.
Not as rosy, but here in Germany I see many job-openings, many remaining vacant - and the hard work done mostly by immigrants. Immigration into our labor-market Dubai-style would be just as great here as in Dubai. As immigration into our social-security is often easier as into our workforce ... well, 'show me the incentives; I show you the outcomes to expect'
I've always sort of assumed, without knowing for sure, that libertarians in favor of open borders aren't actually one-issue but rather have decided that open borders is the only chance we have of getting rid of the Welfare State, since the two cannot coexist, and consider on balance that it's worth the turmoil that would ensue. I'm kind of with them up to that very last step but can't quite make myself go all the way.
If we go that route, we need to get rid of the welfare state first. The problem is that politicians won't want to do that because maintaining the welfare state gets you votes, especially if you bring in new welfare recipients to vote for you.
Sure; that's why people like (I assume) Caplan don't believe you can get rid of the welfare state first. The only way is to make it so untenable that the incentives for the politicians push them in that direction. Like I say, I can't disagree, as we've never had any real success cutting it back even marginally -- how the founders would have laughed at the suggestion that we would spend more on old-age pensions than on the military! I'm just afraid that the downsides of open borders are too terrible to accept even if it allows us to get rid of the welfare state.
That is certainly not Caplan's position. While he is very critical of welfare-stateism, he is ardently pro-open-borders. He even refutes Friedman's thesis (welfare and open-borders: incompatible) in his (great) book "Open borders", Caplan there : We let (our) babies in. Into our welfare-state. Without being sure they will not turn into welfare-trash. Ergo: No fundamental incompatibility.
"If an immigrant plans to live in their new country forever, don't they have the same incentive as any native to ensure that the country they live in stays strong?"
Yes, but an unassimilated immigrant also has the incentive to transform the country they now live in., to something culturally similar to where they came from. These two can conflict, if e.g. the culture of their ancestral homeland wasn't conducive to national strength. And pretty much by definition, their ancestral homeland wasn't strong enough to hold on to them.
Traditionally, most immigrants haven't had the power to do anything really transformative along those lines, and had to settle for maintaining enclaves of their home culture for a generation or two. But as we move into an era with large-scale migration coupled with the attitude that assimilation is problematic because something something settler colonialism, salad bowl rather than melting pot., we should want to watch carefully to make sure that's still the case.
> Traditionally, most immigrants haven't had the power to do anything really transformative along those lines, and had to settle for maintaining enclaves of their home culture for a generation or two
Lots of examples where they have though, like European immigration to the Americas, Australia, New Zealand. Or the Chinese in Malaysia and Singapore.
Right; conquest and immigration are two different things. And *should* be fairly easy to distinguish, but some societies seem almost eager to fuzzify that border and just cross their fingers that immigration doesn't turn into that other thing.
Which is to say, I understand the appeal of Amelia rather more than I'd like to.
>but some societies seem almost eager to fuzzify that border
Notably the government of the UK - which seems intent on at the very least tolerating, and perhaps even encouraging, calls for Sharia to replace British law from Muslim immigrants.
I think Amelia's views go a bit overboard (though I'm writing from the USA, so I see assimilated immigrants, who have adopted American values and customs, as harmless - which differs a bit from the somewhat more insular views Amelia is an avatar of). Though any policy resulting in people like Jihad Al-Shamie being in the UK is a terrible mistake.
I think what strikes me as objectionable regarding that quote is that cultural competition probably isn't going to work very well if the participants don't agree on the rules, the limits of this competition, or even what constitutes winning, and if you put people of disparate enough cultures together, they're inevitably not going to agree on those things. For example, people of different cultures seem to have a wider range of views than might be hoped for on the question of when suicide bombing becomes a legitimate tactic to employ in pursuit of one's political goals. I guess I'm an enemy of freedom if I don't really want to engage in cultural competition with people who have significantly different views on that subject than I do?
Imagine my shame.
I will try to interrogate my anti suicide bombing prejudices.
Move over a bit, Gordon, I need to join you in the Corner of Shame Shame Shame 😁Ooh, do we get special hats to wear? I never got to wear a dunce cap before!
> If a culture of cooperation meets a culture of defection the defecting culture will “outcompete” the cooperating culture.
First of all, 'that's not how this works, that's not how any of this works.' Cultures are by definition insular groups interacting more with each other than with people outside the culture. Cultures don't compete by playing games with each other, so much as by playing games internally and competing on performance metrics. A 'defect culture' mostly means a lot of defectors interacting with each other, which will do worse than a cultures comprised of lots of cooperators interacting with each other.
Second of all, even if it *did* work the way you're modeling here, there's a reason defectbot doesn't win those competitions. The point of cultural competition isn't to have exactly 2 cultures and see which one wins, the point is to *discover tit-for-tat-bot (or whatever the actual optimal strategy is) through *lots* of cultures competing *and mutating*.
> A 'defect culture' mostly means a lot of defectors interacting with each other, which will do worse than a cultures comprised of lots of cooperators interacting with each other.
This is pretty much exactly what we see, no? Certain cultures lead to high degrees of educational attainment, others lead to female genital mutilation and oppression. Cultures that respect individual rights, and have a sense of collective responsibility beat out those without much respect for individual rights, and are mostly every man for himself (at least every clan for themselves).
> The point of cultural competition isn't to have exactly 2 cultures and see which one wins, the point is to *discover tit-for-tat-bot (or whatever the actual optimal strategy is) through *lots* of cultures competing *and mutating*.
A tit-for-tat-bot in this context would be a culture that cooperates, and punishes defectors. I.E. “If you don’t conform to the cultural norms that make us successful, you do not get to participate in our society.” Except that this sort of strategy is prohibited by open borders, since you *can’t* exclude people and cultures that “defect”.
The whole concept of cultural competition + open borders seems obviously wrong, since I don’t see any successful method of exclusion other than “here is a border and you can’t come in.” There can be no tit-for-tat strategy without the possibility of excluding (or harming) the defectors. So a cooperate strategy is a major loser in the long run.
>This is pretty much exactly what we see, no?
...yes, it's just the opposite of what you predicted would happen:
>I’ve seen Bryan Caplan’s view on this stuff, and I’m not sure I 100% understand it, because at face value it seems obviously wrong. If a culture of cooperation meets a culture of defection the defecting culture will “outcompete” the cooperating culture.
>There can be no tit-for-tat strategy without the possibility of excluding (or harming) the defectors.
The issue being that you're inherently framing it in terms of collective punishment for a whole race/nation/etc, and conflating 'culture' with 'race/nation'.
Central to Caplan's idea here is individual punishment rather than collective, and the idea that any race/nation will comprise many different competing cultures, and any individual can decide to join any culture if they find it more appealing than their current one. That's how the competition leads to differing base rates over time.
> ...yes, it's just the opposite of what you predicted would happen:
Then you misunderstand me.
If you look at cultural competition as a process where self-contained cultures are competing with each other, with culture A being full of cooperators and culture B being full of defectors, the cooperators as an entity will do much better. That's not what I mean though. If you look at culture A and B existing within the same society, interacting and mixing, the defectors will make the cooperation strategy ineffective. It will be weighed down by free riders.
> The issue being that you're inherently framing it in terms of collective punishment for a whole race/nation/etc, and conflating 'culture' with 'race/nation'.
I don't think people should be judged collectively. I think they should be judged individually and allowed to immigrate given the intention and capacity to contribute positively to wherever they are immigrating, which are most people in this world.
I simply don't see it as viable for there to be cooperative cultures if there aren't fences around who can join and become a free rider. Caplan seems to frame the issue of immigration as one of freedom of movement, and not one of the strong economic incentive for most of the world to move to high-welfare societies.
If you exclude people from joining your political and welfare system, but have open borders, then you get a large population of people living side by side that are essentially separated castes. The gulf micro-states seem like a good example of this. If you don't exclude people from your political and welfare system in a society with open borders, the flood of free-riders (and their corresponding political power in a democracy) threatens the economic welfare and institutions that made it a desirable place to immigrate in the first place.
> In liberal democracies, with welfare, it seems impossible for the open-borders world Caplan advocates for to function. Population equates to political power, and there is little incentive to *not* game welfare when you have no preexisting social bonds to the society you’re receiving welfare from.
Doesn't this already happen with non-immigrants? Preventing immigrants from accessing these programs (which is already generally the case) seems like it would make this case *easier* to solve than the one where citizens abuse it.
In any event, sometimes culture just changes. Might not have to do with immigration or anything that can be reasonably prevented. Scott's written about this phenomenon before (https://slatestarcodex.com/2016/07/25/how-the-west-was-won/). Having institutions that are resilient or even anti-fragile to cultural change is far more important--and indeed, this is *kind of the whole point of individual freedom and liberal democracy.* Trying to preserve 1776 American culture in amber would be like teaching Newtonian mechanics instead of relativity in the name of science. Allowing culture to adapt is exactly how you overcome issues like losing out to a defect culture, not fighting an impossible battle to suppress opposing culture.
#15 - AI power generation problem - good discussion that “we” collectively will be fine but there’s such a race that the players won’t stop complaining because complaints get your project higher up in the queue
Re 55, I wouldn't necessarily be surprised if property crime had dropped much more then violent crime. I think of property crime as being much more incentive driven and faddish then say assault or murder which more emotionally driven on average. The ubiquity of cameras, security systems, surveillance, theft guards on cars/phones etc. make many types of property crimes riskier and less lucrative (assuming your in an area the police will pay any attention at all).
On the flip side, rather then a change in reporting rates for property crimes, it may just be compositional change from types of property crime more likely to be reported (break into a house / car for ex.) to ones less likely to be reported, or even detected, (shoplifting, porch pirates, that whole Amazon packages on the train thing).
e.g. people who used to break into houses are now porch pirates and shoplifters as the former will get you jail and the later will get you, at worst, a disapproving glance.
#4: It reflects a substantial decline in lab leak theory among the sort of extreme weirds who think prediction markets are the coolest thing ever. And sometimes those weirdos have their predictions anchored somewhat in reality (or in other people's perceptions of reality) by the prospect of making or losing money. But as you point out, it's extremely unlikely that this claim is going to make anyone any money any time soon. So there's no more reason for their predictions to match reality than e.g. a fantasy football league's. And like the typical FFL, the real incentive is to feel like you belong in the tribe.
That lab-leak is declining among the rationalist-adjacent weirds who do the prediction-market thing, is I think driven by the Rootclaim debate and the discussion surrounding it in rationalist circles (like here). We can disagree as to how much people should update on that debate, but I'm not seeing this finding adding anything new.
Are the “suburban neighborhoods on top of giant multi-story buildings” pictures AI slop? There are cars up there. How did the cars get up there? Why? How will they get down?
Same way they get to the top floor of parking garages, I assume.
Cars get to the top floor of parking garages by way of ramps that take up a large fraction of the internal volume of said garages, and which incorporate still more parking space. That's a huge cost and architectural constraint to impose on a large apartment or office building that really wants to have its internal space configured very very differently.
And to what purpose? Yes, the people in the "suburban neighborhood" want to have cars. So do the people in the apartments/offices, so there's ultimately going to have to be parking for all those cars. Probably underground. But in any event, there's no reason for the suburbanites to not park their cars in that same place, and it would cost them hugely to indulge any "but *my* car has to be in *my* garage or it isn't *really* suburbia" fetish.
I would be very surprised if anyone building a quasi-suburban low-density housing project on top of a large apartment or office building, were to incorporate streets and driveways and garages. I would be not even slightly surprised if anyone trying to bash together a quick visualization of a "suburb on top of a highrise" were to lazily copy an existing suburb or suburbs.
You did see that Scott posted the location elsewhere in the thread, right? https://maps.app.goo.gl/c2RN2yGPQbGrfxdz6
There’s a style of apartment building called a “Texas donut” where the building occupies a full block, there are apartments around all the edges, and a parking garage in the middle with whatever sort of ramp goes up it. This is popular because it means that everyone can have their parking space right next to their apartment, and it makes more efficient use of the land area by using the interior part of the block for parking rather than needing to cut out bits of the building envelope to ensure there are windows in the center.
If you just take one of these Texas donuts, and put some townhouses on top, it looks like this.
Also, I think "ramps are also parking spaces" might just be an American thing? I don't think I've ever seen it in Europe. Not allowing parking on the ramps would make them a lot smaller.
> But in any event, there's no reason for the suburbanites to not park their cars in that same place
Because you want your car to be right outside your door, not in a carpark ten minutes walk away. That's just convenience.
Here's a video of what I'm pretty sure is the development, we emerge from the mall to the "suburb" at about the three minute mark:
https://www.youtube.com/watch?v=PAmToIddFuo
> it would cost them hugely to indulge any "but *my* car has to be in *my* garage or it isn't *really* suburbia" fetish
Sure, but these are by local standards ridiculously expensive houses for rich (but not ultra-rich) people, it's okay that it costs them dearly. They are okay with paying extra for the convenience of being able to park right outside their house.
I don't understand why the inconvenience of spending several minutes driving up/down ramps (especially during busy periods) outweighs the inconvenience of walking a short distance to and from an elevator. Or does the building not have enough elevators?
Because you have to do it carrying stuff, or children, or freaking Ikea furniture.
Also there seems to be a special ramp straight from ground level to the roof.
I don't get why people are questioning this. Have they never lived in a place where they don't have their car parked right outside their front door, or have they never lived in a place where they do? Because I gotta tell you that was one of the best things about upgrading from an apartment to a house.
I'm just trying to think about the daily tradeoff. A dedicated 10 story ramp would make sense if elevators are not available or far away, or if power is frequently unreliable. The article seems to focus on avoiding street level flooding as a bonus but in a flood one would be stuck and unable to leave and power likely out. The whole thing sounds dystopian.
Alternatively, they simply took an existing neighborhood, jacked it up, and constructed a big building underneath. The cars are owned by the sort of people who don't read paper notices attached to their doorknobs.
"Mum, what's that ominous rumbling sound?"
"It's barely a four-pointer, go back to sleep."
From Google Maps, this appears to be a similar (maybe the same?) place: https://www.amusingplanet.com/2019/08/jakartas-rooftop-villages.html?m=1
There's a ramp, according to the article.
13. The $170 billion number is just for "23 of the world's poorest countries with high-quality survey data." The estimate is $318 billion per year to end poverty worldwide.
From https://cega.berkeley.edu/end-of-poverty-faq/, "How much would it cost to end extreme poverty worldwide":
New estimates suggest that lifting everyone above the $2.15/day extreme poverty line would cost about $170 billion per year, or 0.16% of global GDP, in the 23 poorest countries studied, which collectively account for around half of the world’s extreme poor.
The analysis then uses the results from these 23 countries, all of which have recent and reliable household survey data, to estimate the cost of ending poverty in countries where such data do not exist or are not publicly available. This extrapolation implies that it would cost $318 billion per year to reduce the global poverty rate to 1%. This amount corresponds to 0.3% of global GDP.
Re #39, about traditional societies sucking pretty bad. For those of us who got a rosy view of cultural evolution via The Secret or our Success, have a good skim of Sick Societies by Robert Edgerton. Traditional suckage is, unfortunately, pretty much the default.
Or check out Jane Psmith's review: https://www.thepsmiths.com/p/review-sick-societies-by-robert-b
Yeah, the Psmiths are a bit of an international treasure! I probably got it from there.
May be of interest to ACX'ers. I recently wrote a critique of "Lizardman's Constant is 4%", relevant to interpreting polling of any kind
https://infovores.substack.com/p/ice-is-not-the-lizardman
Thanks. I saw that, and I'll grant it's possible you're right about the literal lizardman example (or partly right), but I think my overall point still stands and there are examples that are even less defensible than the lizardmen. For example, the 5% of Obama voters who said they thought Obama was the Antichrist, or the mischievous responders in https://pmc.ncbi.nlm.nih.gov/articles/PMC6215371/ . I do think your point that some polls adjust for this carefully is well-taken, and that we may be able to take those polls in particular at face value (although I don't know enough about this to be sure, or to know which polls these are).
Thanks for taking a look! I certainly don't claim that there are never survey results we should dismiss, but I do think the Lizardman Constant can be an impediment to maximizing what we can learn from polling.
When I look at a list like this from Randal Monroe I see a lot of 5% or less results that seem very reasonable to me (surely a nontrivial percentage of people are fine with phones in movie theaters, dislike the olympics, or are less than satisfied with their friends for instance). But several smart ACX fans in the comments are so wedded to the lizardman concept that they are unable to appreciate this.
https://substack.com/@ageofinfovores/note/c-209625073?utm_source=notes-share-action&r=1lmgqf
> the 5% of Obama voters who said they thought Obama was the Antichrist
But in serious theology, "the Antichrist" is not a supernatural hell-spawn trying to take over the world and bring about Armageddon — it refers to any worldly peace-bringer whom the masses come to worship with the kind of fervor they should reserve for the Messiah. It's not inconsistent to vote for a politician while believing him to be "the Antichrist" in this sense, so long as you're voting for him in a pragmatic way without getting taken in by the pseudo-religious hype. I can easily imagine a Christian who votes for Obama due to viewing his platform as the lesser of two evils, but is very concerned at the hero-worship of him and how it detracts from Christian humility, and expresses this as "Obama is an Antichrist".
You may be right! I was also thinking that over the phone it is easy to mishear or misunderstand “the antichrist” as being anti- (or opposed) to Christ. Particularly if some respondents are not very knowledgeable. This is not deliberate tampering with the survey or even really an insincere belief, just a related concept that in an ideal world would be easier to separate out from what they were going for with that particular question. This kind of mistake doesn’t necessarily suggest problems with other questions on the same survey either.
Screw etymologists, *neorxnwang* is obviously just an Old English spelling of *nirvana*.
I noticed the phrase "champing at the bit" showing up much more recently than ever before. Is it one of those things where you learn something new and suddenly it shows up more often because you notice it, or has Scott really been using it more? Or does the AI progress give many more opportunities to use it?
Probably the first. Google ngrams says that both "chomping" and "champing" at the bit have been trending down slightly for the last eight years. Most sources correctly used "champing" until 1999, when "chomping" overtook it in a rise that had started in 1965 or so. "Chomping" is currently about 70% more prevalent than "champing". Alas.
I actually hadn't noticed Scott using either one, but I also didn't just learn the difference. :-)
Nice!
60: Some extra speculative context on cancer and Alzheimer's:
Trade-offs between cancer and degenerative disease crop up fairly often. Both of them are different manifestations of increased cheater cell load in old age.
Young bodies are very good at repairing the body through regeneration. Old bodies tend to have very weak regeneration. There are some reasons to speculate this could be adaptive - weaker old-age regeneration resulting in longer overall lifespans.
Each cell in a multicellular body has an ability to bring down the entire organism by hotwiring its "multiply right now" button to always be switched on. As a result, evolution of multicellularity resulted in all sorts of permission-slip style systems that strictly control when replication is allowed: tumor suppressors.
Young bodies have the same genome and the full set of tumor suppressors in all cells. It's a high-trust society where the risk of having a cheater cell is low. If tissue gets damaged and some cells get lost, it's pretty straightforward to tell surviving cells to proliferate to close down the gap, and expect that they will proliferate in a responsible orderly manner.
Old bodies are riddled with mutated precancerous cells - cells that have some of the locks removed. The term for this state is somatic mosaicism. Somatic evolution pushes precancerous cells to out-compete and out-multiply normal cells, so over time their number is constantly growing. Mosaicism is pervasive in proliferating tissues like skin and intestines, but is also detectable in largely post-mitotic organs like the brain.
So evolution encounters a trade-off between "being healthy" and "surviving longer":
1) The old body could give cells the permission to proliferate and repair the tissue. This keeps the body functional, but at the same time it's removing locks that prevent the precancerous cells from going "all in" and destroying the body by forming tumors.
2) The old body could enter the "coup-proof" curfew state, and prevent all cells from regenerating by triggering cell senescence or tissue inflammation. This "low trust" tissue state protects from cancer, but leads to the frailty and tissue loss common in many diseases of old age.
To massively oversimplify, if you die of cancer at 60, your body chose option 1. If you die of degenerative disease at 80, your body chose option 2. If you live to 110, you were somewhere in the middle, plus were lucky enough to avoid a cancerous cell getting all of its tumor suppressor locks mutated.
This "aging as a tumor suppressor" idea is rearing its head in biogerontology from time to time, usually when the new crop of anti-aging therapeutics fails to deliver results. To be clear, it's far from a consensus view - but to be fair, there's almost nothing in biogerontology that's consensus.
One question is "Why would this regeneration-cancer trade-off apply to Alzheimer's when neurons don't proliferate anyway?" - I'm not sure. I'm not very familiar with brain aging, perhaps someone could expand here. Under "low-trust conditions" lens, brain degeneration could be an unfortunate byproduct of the pro-inflammation endocrine factors clamping down on somatic mosaicism in rapidly proliferating tissues like skin, hematopoietic system, or digestive system. More relevantly to Cystatin C, remodeling of extracellular matrix is another surface for a trade-off: making ECM more rigid in old age helps encapsulate nascent tumors, but also solidifies all sorts of plaques. So you'll see cancers expressing pro-metastasis factors that appear to have "anti-plaque" function. That's like having battle engineers of the invading military force repair bridges that were blown up by the country's own defenders.
Cool concept — it'd be interesting to see an exogenous DNA repair mechanisms that detects and repairs somatic mutations. Or a method to detect less-mutated cells that are in proximity to heavily-mutated cells, then simultaneously inducing apoptosis in the mutated cells and inducing the less-mutated cells to fill in the gaps. (at least, less-mutated along tumorous axes)
The difficult thing about the first approach, assuming you figure out delivery, is figuring out how to not create more mutations than you fix. The tools we have for DNA modification are prone to error, even more so in the uncontrolled environment of a living body, and that problem gets way worse when you're trying to edit oncogenes. 0.01% chance of error * 100,000 cells = whoops, you might've given the patient cancer.
The difficult thing about the second approach, assuming you figure out how to identify heavily-mutated cells and deliver your medicine, is that precancerous cells commonly have mutations that resist apoptosis. That's one of the major functions of cancer-related mutations. If you apply some apoptosis-causing agent + some growth promoter to a population of cells, a small percent of precancerous cells will evade apoptosis and end up exposed to the promoter, and whoops you gave the patient cancer again.
We might see some cool and innovative cancer prevention measures like this eventually. Just emphasizing why this research is so hard.
Would there really be any evolutionary pressure to keep the body alive longer in old age at the cost of increased frailty? I'd have thought in an ancestral environment you're likely to die anyway if you're making either choice, and in any case you're very likely not reproducing anymore. So my naive expectation would be that evolution just programs whatever works for younger bodies that it actually "sees clearly" and it "doesn't think about" what happens to older bodies very much.
You’re effectively describing Medawar's "Selection shadow", and generally, yes, you’re right, evolution won't design an "aging program" specifically for the post-reproductive life stage. But to see cancer-frailty tradeoffs, no new "programs" need to start in old age. The tradeoff can simply be the consequence of tumor suppression strategies that work the same pre-reproduction as they do post-menopause.
What's happening here isn't evolution looking at the 80-year-old human and deciding "Let's make him frail to keep him alive until 90". What probably happened was evolution looking at the 10-year-old proto-mammals who keep dying of carcinoma and deciding "Let's give these creatures a hair-trigger tumor suppressor so that they survive until 20 more often". You can expect lifespans to rise in this way as species shift to ecological niches that reward slow life strategies: memory formation, sociality, caring for the young, predator avoidance, niche construction, basically anything where benefits to your kin compound over organismal time.
To achieve that, evolution selects for aggressive tumor suppression in response to cheater cell burden. Long-lived animals evolved to shut down replication at the slightest signal of DNA mutation. This adaptation is highly useful for a pre-reproductive organism and lets the young survive the cheater cell burden typical for their age. The "trade-off" is just that this same hyper-vigilant safety lockdowns keep happening more and more often as the number of mutated cheater cells rises steadily over the decades.
So under this lens, evolution didn't "program frailty" for the elderly. It programmed "high-security anti-cancer lockdowns" for the young, with side effects that ramp up over time until they're just as deadly as cancer itself.
(As an addition, there are some good reasons to believe post-reproductive organisms can also be visible to evolution via kin selection - see grandmother hypothesis - but this is not strictly necessary for the explanation)
Re: #55 - Does anybody have a link to a steelman on the claim the murder/general crime rate actually has not gone down? I'm interesting in seeing a good version of this argument - the simple version (gunshot wound care is so good now it's just hard to kill people) falls flat on its face for the reasons Scott mentioned in the post. It also defies casual observation for me - you'd never hear about people going to much of Oakland and neighborhoods like Third Ward, Houston that didn't live there when I was growing up and now they seem mostly fine? And despite living in a kind of crappy part of Houston, not exactly a low crime city, for many years I know don't know a single person who has even gotten so much as held up for their wallet? N=1 of course, but claims about omnipresent danger in American society are so common online it makes me wonder if I'm not just weirdly lucky.
I'd like to see that particular steelman too. Wouldn't have to be all or nothing e.g. a solid steelman for current reductions just being seriously exaggerated, would be worth having.
The reported/gathered drops in US crime rates starting with 2023 have been _so_ drastic -- and so broad, and showing no signs of stopping -- that it's hard _not_ to wonder whether there isn't some big data confounder that's being overlooked. I say that even though as a lifelong big-city dweller my personal eye test is similar to yours.
For example Jeff Asher's stuff on Substack, which appears to be high-quality and thoughtfully done, is lately making my brain go "naaaah, couldn't be. You gotta be missing something."
Holding people up for their wallet is much less useful as a crime these days! A lot of people carry little or no cash, and credit cards and phones are easier to disable remotely than they used to be.
Meanwhile, online fraud is much easier and more lucrative than it used to be, particularly if you do just a little bit more research into how crypto works than the average gullible person.
Certainly, and that may be a driver of why we see less petty muggings in the data - the value proposition just isn't there anymore!
As far as your (potentially not totally serious?) suggestion that these people may have gone into crypto scams - I can tell you grew up in a nice neighborhood haha. The kind of people who'd jack my stuff walking home from school as a kid are not going to understand how to rugpull people in a crypto scam. *Maybe* some simpler online scams, but I'd also question how much those really happen. Even back when the US was much poorer, you had to be truly dumb to have to rely on petty crime to get by. Criminals smart enough to figure out how to spoof an email and convincingly clone an e-commercd sites
UI were never the guys sticking a knife in your face for a wallet in some counterfactual where the Internet doesn't exist. They'd be involved in some kind of white collar crime or organized gang that consciously avoids doing things like sticking knives in people's faces, especially completely random people.
Good points! My main point was just that there are some types of crime that have likely gotten easier as ways to make money, even as others have gotten harder. I don’t know which ones would have real substitutions effects. But I wasn’t thinking of a crypto rug pull where you sell high after getting others in - just the kind of crime where you get someone to send you their wallet info or wire you some coins by pretending to be a Russian bride or pretending to sell them some drugs.
Oh sure, that makes sense. I wasn't thinking of those. Yes, those kinds of crimes probably substitute much better in terms of simplicity, risk, and reward (although from a social outcome POV, I'd much rather street criminals scam people via lying in emails than actively threaten physical harm)
For #32, CBS claimed it wasn't suppressed but had enough fact checking issues that they needed to postpone it for verification (and they were always going to release it eventually).
Ymmv on how trustworthy you find CBS, but from the outside view - if you assume an honest non-politically motivated newscaster management exists (even an intermittently honest one), you should expect to see stories like this happening (since individual reporters who lean left will occasionally (a) make a lot of trump-critical reports with twisted facts and (b) claim to be suppressed in the face of fair pushback).
I’m no expert on clinical trials but I wonder whether there’s a dynamic somewhat similar to what I see in environmental reviews of infrastructure projects. A lot of money is at stake and project sponsors are extremely risk-averse, so the analysis often goes far beyond what is actually required by the applicable regulations in an effort to make the process bulletproof against lawsuits (even though NIMBYs often sue anyway just to delay and make the project more expensive).
I am an expert on clinical trials. This is one of the dynamics at play, but not the only one. I'm planning to write a more thorough response under the linked article. It has some good insights, but also some of the things he mentions are not right at all.
I spent a couple years programming software for data collection to support clinical trials. Running on Palm Pilots - to give you an idea of how long ago. What the linked article doesn't mention is that electronic data collection is much better than paper-based because it avoids what they call the "parking lot problem". The subjects fill in the last months worth of questions just before they turn in the papers rather than when they were supposed to fill them out - the electronic method forces them to answer the questions in a timely manner. I would have expected this superior methodology to win out over time, but now I'm not sure.
Ah yes, "parking lot compliance". I've had coordinators tell me how they sometimes watch patients sitting in their car filling out the patient diary.
Nice thing about ediaries is that they remind the patient when they should fill them out. Problem is older patients don't understand them.
The diary data is usually an exploratory endpoint (since it's unreliable), so nobody wants to spend the extra money on them. Also paper seems like it should be cheaper (before the extra SDV time). Sometimes we pay for them if they're important to the primary endpoint, but usually they're not so it's a hard sell to spend extra money on a vendor in order to save money on data entry and verification.
> Oh well, I’m sure tech journalists are right about everything else being a giant Ponzi scheme that will inevitably collapse and reveal the entire tech industry to be an emperor without clothes.
You picked a hell of a day to say this.
(Context: Bitcoin crashing at an insane rate this morning, to the point people are talking about the crypto bubble popping)
Banned for low-effort high-temperature comment (I wouldn't have banned if not for the last sentence).
I am actually happy to see more bans of low-effort high-temperature comments.
I agree that it's a bit atypical, but Scott also engages more with politicized discussion in this post than usual. And I don't think the bans are outside the usual policies, it's probably just that Scott happens to have more time/energy/engagement today.
Yep, same here.
AI is a scam? Dude, it's your problem that you can't use it. For me, it generates websites and translates books, at the cost of €20 a month. But if you think it is a scam, simply ignore it. I know people who refused learning to use computers 30 years ago, because they believed that it was a fad that would soon pass.
(And the Bitcoin "crashing at an insane rate" is still 20x as expensive as when I bought it, and I expect it to stay that way.)
If this debate is supposed to be "rationality-adjacent", at least the participants should not clearly contradict reality and be smug about it.
Cursory search suggests this might be relevant to #55. In-hospital mortality from GSW has fallen, but pre-hospital mortaility has increased so in net GSW mortality hasn't fallen much.
https://www.thetrace.org/newsletter/analysis-shooting-victims-gun-violence-intensity-trauma-baltimore/
I appreciate that you used "champing at the bit" in this post. Esoteric traditional spellings huzzah!
I got it wrong originally and someone in the comments corrected me.
I've been chimping (out) at the bit about this comment....
Not very consequential, but funny: The super-secret Coca Cola formula has apparently been duplicated, with some mass spectrometry and a lot of testing. Since it was never patented to avoid publishing it, anyone can now make something indistinguishable from Coke. You don't even need coca leaves. https://www.youtube.com/watch?v=TDkH3EbWTYc
Regarding 21 apparently, we already have a timecapsuledLlm only trained on data between 1818 75, although it’s apparently a small model as you might expect. Unfortunately, I can’t find any copies of it online, which are directly usable, although that’s apparently in part because I’m not a programmer since Grok assures me that a normal programmer can easily use the publicly available data to run a copy of this. Still, it does seem like that kind of thing you can use for interesting experiments and trying to find out how people would have justified their world view back in the 1870s, which does strike me as super interesting. Actually, if anyone has a clue, how one might go about turning what is publicly available into something I can prompt. I would be extremely grateful because right now grok’s explanations while pretty confident leave with even more questions as it keeps using and talking about concepts, I have never heard of. Maybe I should try asking ChatGPT or Claude instead.
What are we doing here with the AI forecasts? In 3-5 years will it be pushed back 3-5 more years? This whole exercise feels unhelpful, why not just admit that everyone is just guessing, a fact we knew in 2012? https://www.lesswrong.com/posts/47ci9ixyEbGKWENwR/ai-timeline-predictions-are-we-getting-better
I don't think your link supports your point - it says most forecasts cluster 20 years out, but now the forecasts are 2 - 10 years out, presumably because AGI is actually getting sooner. Over the past few years, most people's timelines have gotten shorter rather than longer - see https://agi.goodheartlabs.com/ .
Is naming your AGI forecasting project to evoke Goodhart some kind of nominative determinism joke?
> #22, New representation-in-historical-movies controversy, this time about an African woman getting cast as Helen of Troy in the new Odyssey...
Funnier still: my friends who are expat Greeks are now asking the question of why many of the major actors don't look Greek or of Greek extraction, with Matt Damon being the most obvious Vóreioi in the film.
I think it's fair to say that there's a spectrum of Greek-passingness, with Jason Mantzoukas at one end, then Omar Sharif, then Orlando Bloom, then Matt Damon, then Zendaya Coleman, then Lucy Liu, and then finally Nupita Nyongo who is just about the least Greek-passing person on Earth.
What about Anthony Quinn as Zorba the Greek?
I've never seen the movie, but I looked up some pictures. He doesn't look totally implausible, but his face looks more Greek than his hair. Somewhere on a par with Sharif.
:-)
It certainly bugs me when historical dramas or dramas set in supposedly historical realities use actors with obviously non-period-authentic ethnicities in their cast. I found Bridgerton particularly discordant, to the point that I couldn't watch it. Primarily because it whitewashes the actual existence of slavery in that period. I don't have a problem with mixing ethnicities and sexes in Shakespearean plays, because even his history plays are mostly history-free. As for Homer, it's mythical enough for me to be to *mostly* unbothered by the casting, but it does seem that there should have been some ethnic consistency within tribes/kingdoms as Deiseach said below. Lupita Nyong'o would've made a fantastic goddess, IMHO—Athena or Circe.
I think Lupita Nyong'o would be fine in a part like, say, Calypso but not for Helen of Troy. Unless you're going to cast *all* the Spartans/Laconians as black Africans, which would include Menelaus and Agamemnon, which I could live with as at least it would be consistent and a different, interesting choice about how these are all representatives of separate Greek kingdoms and not some unified state called "Greece" that did not exist in that time:
https://en.wikipedia.org/wiki/Helen_of_Troy#/media/File:Homeric_Greece-en.svg
That would be Diverse Casting that works! Different kingdoms are internally consistent but the Spartans don't look like the Macedonians don't look like the Thracians and so on.
Plus, Nolan's costume and set designs seem to be deliberately trying to make everyone look as unattractive as possible, so we're in the "mud and earth tones and black leather" past of TV and movies, and not the "people liked colour and gold ornaments" past.
>Different kingdoms are internally consistent
Is this a message that Hollywood is interested in sending, though?
> Different kingdoms are internally consistent
That would be an interesting casting challenge. Out of curiosity, I did a cursory search to see if there was a substantial genetic change in modern Greek populations from their Bronze Age predecessors, and Greece is one of those rare places in Europe where there is substantial genetic persistence. Usually, conquerors and migrations shuffle up the genetic deimcks, but, depending on the region, populations are 70 to 90 percent similar to their ancestors (at least that's the claim made in this popular article). And overall Greeks from the islands are genetically similar to mainland Greeks.
https://greekherald.com.au/culture/history/new-dna-study-shows-todays-greeks-similar-those-2000-bc/
>36: Indonesia has solved the conflict between density and single-family zoning by putting suburban neighborhoods on top of giant multi-story buildings (h/t @xathrya):
How many Indonesian kids need to die chasing their soccer ball off the top of a skyscraper before we start to care?
Also, who would call the depicted neighborhoods "suburban"? I live in a pretty dense suburb on the SF Peninsula, in a lot that is smaller than any of my neighbors', and I have more of a yard than that.
Perhaps "superurban" ("above the city") is more appropriate.
Heh. But seriously, didn’t it look more like the cheek-by-jowl townhouses of a city than like anything you’d call suburbia? Doesn’t mean it’s unlivable, but it’s a very different niche.
It seems like walls are a cheap and effective solution to that problem.
Do kids actually walk off cliffs?
Only if all their friends do.
Surely the relevance of no 13 is that we are spending that much _every two years_ and have not meaningfully reduced absolute poverty in the countries we are spending the bulk of that money in?
But it is a bit of a vicious (or only partly virtuous) circle in that currently our aid money only makes a minor impact on demand for Harrods' shopping, custom made suits, expensive cars and watches, expensive hotels and restaurants in the local city center, and private jets. Which, apart from the hotels and restaurants, only modestly impacts inflation in these markets and only outside of the countries we are giving aid to.
Direct cash transfers on that scale in such dysfunctional economies would likely trigger substantial (NGDP growth and therefore) inflation in the local economies, which would raise the absolute poverty bar since people could no longer afford to subsist on the original amount.
But course we are spending this money _every two years_, and direct transfers would substantially jncrease demand for good governance, this would absolutely work to massively reduce absolute poverty even despite a few localised bloodbaths.
> Direct cash transfers on that scale in such dysfunctional economies would likely trigger substantial (NGDP growth and therefore) inflation in the local economies
That would only be true if they can’t get imports. If they can get imports, then the cash transfers end up turning into imports (though there might be some inflation in the cost of locally-supplied services).
That's the point though. Leaving aside that a _lot_ of the demand would be for construction and personal services, indofar as goods are concerned these countries can't quickly get any large amount of imports, their economies are not functional enough to facilitate that.
This could ofc be alleviated by ramping up the transfers over a few years but even then I predict substantial inflation.
51: When I wrote my post on using light glasses to treat non-24 (and that they can treat SAD), I specifically looked for any mentions in the ratsphere about them, and found none. This is confusing, because they are actually a viable alternative to lumenators! They cost the same or less, are portable, and don’t change in effectiveness based on gaze direction. They’re strong enough to entrain my circadian rhythm, and they have similar melanopic lux to lumenators. I want more people try them out and report on them!
Light glasses (I use Luminette): https://optimizeyourbiology.com/best-light-therapy-glasses
My post on treating non-24: https://www.lesswrong.com/posts/mHJFu6FAJc4ikscnq/like-night-and-day-light-glasses-and-dark-therapy-can-treat
Hey Drossophilia. I'm also on team non-24, but I've gone into 'remission' (if that's the right term) as an accidental side effect of taking a low dose of Escitalopram. It was weird, in bed every night at ten. That lasted about six months before slipping back into delayed sleep phase, where it has stayed locked for a few years so far. I haven't found any studies that mention this effect, but I thought I'd mention it to you. Cheers
Oh wow! I have a somewhat similar effect taking bupropion and memantine (is what I think is happening). Unfortunately it makes me only sleep for 6 hours, so I’m quitting the latter. Thank you for sharing your anecdote!
The theory I have right now is that non 24 is like ADHD for sleep. One of the major side effects of Escitalopram is apparently 'insomnia' during the first few months. So maybe just as stimulants are a good treatment for ADHD, perhaps 'insomiagenics' (to coin a term) might be effective treatments for non 24? No idea really.
I may have to add a note about that in my post! That’s a really interesting way to think about it…
I haven't studied Soviet history, but just going off Sarah Paine's lectures here :) I assume we have to keep the extractive nature of the whole enterprise in mind, or else the numbers feel backwards. Like if we lost "Calikraine" with 18% of our population and 16.2% of our GDP, sure that would be a loss in terms of total national power and prestige, but wouldn't our relative GDP-per-capita go *up*? Well maybe in Loose-US-Analogy-istan it would, but not if Moscow (Moshington DC? New Mork?) had been systematically siphoning wealth from the provinces.
Relative GDP per capita is a thing, but it's not the only thing. There's also absolute GDP, and there are some people who care about absolute GDP more than they care about GDP per capita. That's a particularly common mindset who expect to be at the top of the pyramid, with all of that GDP being channeled through them.
If you had a choice between being the founder/owner of a billion-dollar-a-year corporation with ten thousand employees (GCP per capita: $100,000) or a million-dollar-a-year startup with five people (GCP per capita: $200.,000), which would you choose? Which would you expect most other people to choose?
> champing at the bit
oh man i almost complained about "chomping" at the bit last post, i'm glad to see someone else got to you <3
Fascinating. Question for the mathematicians out there with respect to #4. Suppose we have an ideal betting market wager which can only resolve in one direction within a finite amount of time (e.g. this can only be proven true, but if it's false, it will remain open forever). What sorts of dynamics would we expect from such a wager, absent any new evidence? (My intuition is we would expect something like exponential decay towards 0% as we approach the limit of infinite time).
If decay is the expected behavior, what interpretations can we make about the wager given the shape of the curve? What would determine the underlying decay rate? Would we expect the decay rate to change given new evidence?
This is a side of better markets I've never thought about before.
We cant make assumptions about the what the shape of the curve should look like without making assumptions about how people's beliefs update in absense of new evidence. There probably is be an assumption of at least one side that by xyz date, evidence should have come out, and if it doesn't then that is evidence going the other way, but we dont know magnitude or direction here.
I think its pretty dumb to have infinite horizon predicition markets, they should just do long horizon instead.
> There probably is be an assumption of at least one side that by xyz date, evidence should have come out, and if it doesn't then that is evidence going the other way
What if the question is about the Second Coming of Christ?
I'm not sure if you're deliberately referencing the article about Polymarket's Second Coming market (https://www.lesswrong.com/posts/LBC2TnHK8cZAimdWF/will-jesus-christ-return-in-an-election-year), but if not, I'd recommend the read.
I wasn't; it's just an obvious counterexample to the idea that there should be "an assumption of at least one side that by xyz date, evidence should have come out".
I am interested in the article, but there's something I'm not following. The time crunch theory is described like so:
> right now, there's not much interesting stuff happening on Polymarket. People are spending a lot of money betting on sports, but not much else. But at some point in 2025, other markets will get a lot of attention. [...] Right now, all these markets have mere single-digit millions of dollars in trading volume, but that could very easily change.
> And if it changes, some of the people betting No on Christ's return will want to unlock that money -- that is, sell their "No" shares -- so that they can use it to bet on other markets. If enough people want to sell their "No" shares, the "Yes" holders may be able to sell out at an elevated price, like 6%, potentially getting a 2x return on their investment!
I'm willing to stipulate the following:
1. Christ will not return in 2025;
2. Everyone is aware of this;
3. The market will resolve NO on January 1, 2026.
I agree that the Second Coming market is a terrible position to hold because it gives you low returns over an unreasonably long investment period.
And I also agree that at some point some NO bettors are going to realize that and sell their NOs, and this will drive down the price of NO.
But the quote I just pulled says that this will somehow mechanically raise the price bid for YES. I don't see why that should be the case. The hypothesis is that a lot of people realize they have more NO than they want, and sell it. This will mechanically lead to a lot of other people buying NOs at lower prices.
But there's no rule that says that when NO is trading at $0.61, YES has to be trading at $0.39. The market could reflect a 61% chance of no Second Coming and a simultaneous 2% chance of at least one Second Coming. The fact that that totals only 63% isn't a problem; we've already stipulated that the true probabilities are 0% YES and 100% NO, and 2% YES / 61% NO isn't epistemically worse than 39% YES / 61% NO - it's epistemically better!
Where would the increased demand for YES come from, in this theory?
Fair point. I'm pretty sure that in practice the probabilities always sum to 100 - there's always someone on the other end of the bet - but I don't think I understand Polymarket well enough to properly grasp why that is.
I've found the relevant mechanism: https://docs.polymarket.com/developers/CTF/merge
The fundamental working of Polymarket is that you can pay $1 (measured in a USD-tied stablecoin) to receive a matched pair of YES/NO tokens.
In "ordinary use", one of those tokens will be redeemable for $1 (of stablecoin) when the market resolves, and the other one won't.
Turns out Polymarket will allow you to redeem a matched pair at any time, whether the market has resolved or not. Which means that if you're strapped for cash and want to sell your 100%-chance-of-NO for $0.94, Polymarket's system can match you against someone who wants to buy that NO for $0.94...
...but it can also match you against someone who wants to sell a YES for $0.06. In that case, you buy their YES for $0.06 and redeem the YES/NO pair for $1, netting $0.94. This is why depressing the price of NO automatically raises the bid for YES.
35: the "mental health" vs "mood" thing may have explanations that aren't merely stigma per se. By default if I saw these on a questionnaire I would see them very differently.
The brand new 2025 Honda Civic you bought is functioning perfectly, but it does not feel as good to drive the Civic as it would feel to drive a brand new Corvette. There is nothing wrong with the health of that Civic, a mechanic can do nothing to make it feel like a Corvette, the Civic is high health + low mood. Maybe liberals' problems tend to be more defects that are fixable, or maybe ~10% of them tend to think everything is supposed to be awesome and perfection is attainable and why aren't we constantly working to make everything sunshine and rainbows for everybody and if it's not sunshine and rainbows for me then there's a defect somewhere. And maybe that ~10% of conservatives are just more likely to accept that sometimes live hands you a Civic instead of a Corvette, it's not gonna magically become a Corvette, none of us like the Civic but it runs as intended, that's life so get in the Civic go to work and shut up.
36: I feel like I've read discussions about putting neighborhoods on top of dense buildings and there's some practical reason you aren't supposed to do this. Sewer lines or water pressure, fire hazard, windbreaks, carbon monoxide from heaters below you, something about heat radiating off the surface, I don't know but intuitively it seems like there's some obvious flaw in this plan.
Re: #12, I think a string use case for AI is answering real life halachic questions on demand, replacing texting a rabbi or searching through long Q&As for something similar to your question. I propose to create an AI benchmark for that. And it shall be called... BenchBench. :)
I originally read FAI-C as FAFC-- as in, f around and find christ.
17 - the acronym doesn't quite work out because of the I in Intelligence
I think 'women asking men out' has a strong cultural valence too. When I (a New England boy) went to college in California, I was astonished to find California girls coming up to me and asking me out. Totally unprecedented, unheard-of behavior in much more traditional New England (and I think the rest of the country too). I had no idea how to handle it either- often I wasn't interested but it was awkward and difficult to let them down gently
The historic German Reich (I.e. the Holy Roman Empire) included Bohemia and Moravia. The cities of Prague and Brno had historically been majority German speaking until the mid nineteenth century and were still regarded as centers of German culture even as late as the 1930s (even if a lot of that “German” language culture was being produced by Czech Jews). Notably, Germany never tried to incorporate Slovakia, which also had German minorities but had historically been Hungary, not part of the Reich. Maybe that had some effect on Hitler’s thinking.
#13: I think this just fails a very basic intuitive test: the flipside of poverty as a state of low levels of consumption is low levels of production, so trying to fix the consumption part via mass fiscal transfers without having any plan to fix the production side is not a recipe for "ending" anything at all. At least Jeffrey Sachs had a theory that addressed these pretty basic elements, even if he turned out to have mostly been wrong.
#27: those are all very nice, but there's a suburban mid-rise apartment complex going up by me with an exterior that mixes four different colors, so I'm certainly not sweating it, India.
#32: I don't have any inside info here, but I think it's likely that your characterization is inaccurate and that this was really an attempt by Blob journalists at CBS to discredit their non-Blob boss, Bari Weiss, for the crime of interfering with Blob output in a formerly Blob-owned factory.
Re 13: lack of capital is one of the major limits on production. If you can buy some sheet metal to set up sturdier walls and roof, then you can spend less of your time doing repairs and more doing services or labor for others. If you can get effective treatment for illnesses then you spend less time out of education and work for sickness. If you get a butter churn or sewing supplies, you can spend a bit of labor to upgrade the raw goods you produce so that they are more valuable for others and sell them for more profit.
Agreed, which is why "just give everyone money" seems unlikely to work. You don't know what portion the recipients are likely to spend on capital goods. The fact that these people are living in extreme poverty in the first place suggests they probably don't have access to many capital goods at any price. Where they live, you can't just go down to the farm supply store and buy a butter churn; that store doesn't exist within any reasonable distance.
Huh, I would have thought the opposite! If you send a lot of dollars to rural villages, then people in developing cities now have an incentive to set up a farm supply store in the rural area, as a way of getting dollars. This lets the market create the supply chains that actually bring poor people into the world of capital.
Try it out. Let us know how it goes.
I do: https://www.givedirectly.org/
Does it work?
"This insurance coverage mandate isn't sized to the typical claim" is, on its face, not a terribly convincing argument given *what insurance is*.
I agree - and if the claims are really that low than $1 million insurance coverage shouldn't be expensive, right?
It's possible that higher coverage induces more aggressive litigation and so claims experience is not independent of coverage. But that's a separate case to make.
(Also, I've been told by others that Empower's driver base is essentially just "drivers who got kicked off of Uber and Lyft", which, if true, makes me pretty sympathetic to DC here.)
If you call one of the big personal injury firms, they’re gonna route you through a pre-screening where a newly-minted lawyer will make a quick estimate of what he thinks your claim is worth. Part of that math involves assumptions about typical coverages, and obviously deep pockets increases the estimates, if you think the at-fault driver was in the course of employment at the time this would raise it of course. I imagine if industry standard went up 2x you’d find all the component parts of a claim drift up to fill the difference, much as US college tuition rose to whatever level of government loans were offered.
Then where does it stop? Why not $10 million? Why not 100?
How about, "how much damage can one realistically do with a car"? The number is self-evidently much higher than $100k but lower than $100MM.
53: "Futurist Cooking, or, the J. Kenji Lopez-Alt-Right?
#36: During World War II Boeing built an entire (fake) 26-acre suburban neighborhood on top of its Seattle bomber plant to conceal it if there were Japanese air raids. The “houses” were only about six feet tall, as bomber crew high above would not be able to tell the difference.
And they hired a Hollywood set designer to do it....neat story that I hadn't heard of before. Some pics here:
https://rarehistoricalphotos.com/boeing-fake-rooftop-town/
Shoulda just printed a very big picture.
> if trauma care has gotten twice as good, then the apparent number of murders will halve even for the same amount of crime
Does this not assume that all murder victims somehow end up in an ICU? I would assume that in the vast majority of murders, the victim is already dead before anybody knows about it, much less is able to intervene.
If only 20% of murder victims actually end up in the hospital at all, then a 50% reduction in attempted murder victims whom trauma care fails to save becomes only a 10% reduction in murders. Maybe my intuitions are wrong here and more murder victims than I expected end up receiving hospital care before they die.
This comment (https://www.astralcodexten.com/p/links-for-february-2026/comment/210473488) has some discussion on this.
Yeah, it seems like for both murders and accident fatalities, improved medical care/improved EMS has to be decreasing the numbers over time, but probably not by a huge jump at any one time.
One interesting thing to do would be to look for some natural experiments here. Like, when an ER opens/closes in the area, do we see the result in the local murder rate/traffic fatality rate?
The source for the $40 billion savings number from the eye medication study says:
> The use of bevacizumab from 2008 to 2015 resulted in an estimated savings of $17.3 billion, which corresponded to a $13.8 billion savings to Medicare and a $3.5 billion savings to patients. This amount underestimated the actual cost savings to Medicare providers, since approximately 30% of Medicare-eligible recipients received care within Medicare Advantage plans and were not included in this analysis. ... Additional savings over the $17.3 billion would have accrued from the use of bevacizumab if diagnostic categories such as diabetic macular edema and retinal vein occlusion were included in this study.
https://pubmed.ncbi.nlm.nih.gov/29655642/
The number for the "additional savings" is quoted elsewhere as in excess of $30 billion, presumably that got rounded up to $40 billion. But importantly this is savings over 7 years, so keep that in mind when comparing to the *annual* budget of the NIH, and that a good chunk of the savings is for the patients not Medicare.
I want to thank Amelia for briefly turning the Culture Wars from something relentless and insufferable into something f@#%ing badass. Just look at this thing: https://m.youtube.com/watch?v=3B4524ot5BM
Agreed!
AI slop.
Ewww.
It's so sad. If the Amelia thing had happened a few years ago, some gifted right-wing troll would have created amazing videos out of it, perhaps by doing Amelia-themed alt-right karaoke of recognizable songs, as Walt Bismarck used to do with Disney songs back in the day.
Whereas all we get now is slop.
Slop that, granted, one day will start to feel truly human, and that will be an even sadder day, because it will mean humanity has been replaced.
The alt-right has it wrong. It's not immigrants that are stealing our jobs. The real great replacement, the real genocide, is the one robots are doing to us.
Oh look, the humans are putting their smartest people on the job of inventing their successor species....
Thanks very much for the shout-out, your writing on kabbalah is some of my very favorite ever of all time ever so it is particularly tickling to know you enjoyed (and hated?) mine. May your vessels never shatter
53: It looks like the link to the Futurist cookbook is broken
Your rejection of critiques ofthe financial status of Uber sure looks identical to a leftist indictment of Uber
elaborate
56. (NIH Funding) - Is it possible that there is actually a useful motivation behind this change?
If NIH is mortgaging their grants, then that makes them sensitive to the NIH budget in future years, which is highly uncertain given the current political climate. So, NIH could end up funding the first year of a 4 year project and then needing to cancel it in a future year due to lack of funds. That would then waste 1 year of resources. If they fund the full grant up front, then the money is already effectively spent, thereby reducing uncertainty.
NSF made similar changes about 15 years ago during another period of economic uncertainty, where they deliberately made fewer grants for a couple of years but spent all the money up front in order to reduce their "mortgage" rates. Their intent in doing so was to reduce uncertainty in the budget.
I don't know the details of this NIH change, but it is possibly not nefarious. Of course, this administration has given us little reason to give them the benefit of a doubt either.
Re 52, Amelia (albeit duplicating what I wrote in the hidden thread - 'scuse, I visited there before seeing this post):
It is ironic, given the (valid!) concerns about "superpersuasion" via AI, and the potential for this use to be part of oppression by governments, that the last few weeks have seen a use of AI video generation by the _resistance_ in the UK. It is doubly ironic that the "Amelia" character was originally created by the Starmer government as part of the "Pathways" propaganda game, to disparage unWoke thought. Instead, her widely shared unWoke views lead to her being incorporated in AI-generated video clip after video clip (e.g. https://www.youtube.com/watch?v=xIdSect2cyo&list=RDxIdSect2cyo&start_radio=1 ) and now she is the symbol and avatar of _dissent_ from a government that has arrested 12,000 people a year for speech.
I don't expect this to last - few governments are _quite_ so stupid as Keir Starmer's, but it is a refreshing irony to see this tool used, however briefly, for free dissent.
It's not uncommon for the right, particularly the alt-right, to adopt and reverse media intended to attack them. One of the big weaknesses of the right is that it doesn't really have a pipeline to produce friendly fiction, because the creative professions in the West are disproportionately SJWs - they're not 100%, but they're enough to keep the rest in line via the threat of blacklisting.
The alt-right's solution is to take SJ-promoting media and simply identify with and promote the villains, particularly when that media fucks up its messaging such that normal people find the villains sympathetic.
"Keep Your Rifle By Your Side" is another prominent example.
Many Thanks! It is amazing how successful it has been in this case - and how much it (currently!) differs from the typical AI-as-a-tool-of-tyrants concern.
It seems sort of predictable to me. Democratizing the means of cultural production seems like it would always lead to more dissenting voices being heard.
Many Thanks! But that the introduction of AI video generation wound up democratizing cultural production rather than being differentially available to better funded and/or more technically sophisticated groups, _centralizing_ cultural production, was not widely expected.
Why wouldn't it be expected? The first time I saw Sora I thought "this is the end for video production studios". Making a video like that before LLMs would've cost, I dunno, tens of thousands of dollars? Now it's basically free. How could that *not* lead to democratization?
Many Thanks! That the price drop was as large as it wound up being was not the only possibility before it actually happened. Consider what would have happened if the technology had followed an alternate path where the quality went up (say to routinely photorealistic quality) but the price had dropped only slightly or not at all.
There are additional hazards that might lead to centralization of control that haven't happened yet but might still happen: Consider that there are relatively few frontier AI labs, so they are potential choke points for censorship. As is, there are some restrictions on content that AFAIK, against e.g. CSAM, that are enforced by that set of labs. If, e.g. the UK government decided to ban 'politically incorrect' video generation, they just have to coerce a handful of labs to enforce it. And detecting which prompts fall into that category would use the _same_ language understanding capabilities that make it possible to generate video from a text prompt in the first place.
I'm thankful that democratization happened, and that dissenters have a voice, in fact aided by the technology. I'm not sanguine that this was the only path that could have been followed, nor that it will persist over the long term.
>the Christian community wants AI that supports our value system and wellbeing.”
Is the Christian God a Basilisk?
Re: link 51 and GetBrighter, speaking as someone who's potentially interested in brighter illumination sources, the price is a sticking point for me; I'd rather pay less than buy something whose materials are selected for an "ultra-premium feel," and it strikes me as highly relevant that I could get the same amount of passive illumination for cheaper by buying a power strip and a set of conventional sunlamps to plug into it.
I already bought a sunlamp a while ago, have been thinking I'd be interesting in upgrading, but checking out GetBrighter just convinced me that the logical upgrade option would be "more sunlamps."
You can see here[0] for how it compares to alternatives and even DIY'ing, we're pretty cost competitive on lumens/$, especially when taking CRI into account.
[0]https://www.reddit.com/r/Lighting/comments/1po7sxb/are_there_diy_alternatives_that_compare_favorably/
It doesn't look like this is comparing to a very wide selection of alternatives- you say that you're "pretty cost competitive on lumens/$, especially when taking CRI into account," but all the alternatives you actually compare on a lumens/$ basis are ones specifically marketed as having comparably high CRI, and while all of them are significantly cheaper than GetBrighter, you can get significantly cheaper than that if you're not looking for options that specifically aim to maximize CRI.
For me, smart home integration has zero value add (if anything, I'd prefer to avoid it,) and I'm not especially fussed over the difference between 85 and 95+ CRI, and with those priorities, I can get similar levels of illumination at roughly a quarter of the price.
Sorry I phrased it poorly. I meant to say we're not too far off lumens/$ vs DIY'ing given we're an actual consumer product (& the features like adjustable cct, dimming, smart home + manual controls). But depends on your threshold for "too far off"...
That's fair. I may simply not be part of the target market for this sort of product, but for what it's worth, as someone who *is* quite interested in high-intensity lighting options, I'm sometimes willing to pay more for better form and convenience, and the higher CRI has at least some positive value to me, but premium materials matter very little beyond the point where I'm confident it won't be accidentally broken, and the smart home connectivity doesn't offer any positive value. I'd be willing to pay up to about half the price of the GetBrighter lamp for something that offers illumination of several sunlamps, shares their adjustability options, but has the added convenience of being a single point source which is convenient to move around the room and doesn't require a power strip or multiple outlets. I have no idea whether exploring a cheaper option without smart home functionality or anything like that is something you'd want to explore, but I suspect there's at least some available market segment to be captured that way.
> Aren’t many murders in very fast attacks where the murderer doesn’t get to choose how many shots/stabs to land?) and would welcome more research on this topic.
I would assume, until I had good evidence to the contrary, that the medical care effect is just not very large, at least not over the timescales referenced here. Have we really gotten significantly better at treating gunshots in the past few decades? (Also, this seems like maybe another way to test the theory--look at murder rate by method, and see if the more treatable ones have seen a bigger decline).
I don’t have any stats handy to prove this, but a pretty large number of murders are dumb impulsive crimes, people getting into a fight and one grabbing a knife or somebody getting kicked out of a party and coming back with a gun. A lot of of those murderers are intoxicated, angry, and not thinking very clearly.
RE 42 (evil delivery report was AI): Couldn't the whistleblower have run his report through an LLM in order to disguise his style of writing?
Yes, I think the evidence that he didn't do that was that when the journalist asked for more evidence he tried to fake it.
For link number one, a bit of a nitpick but meta was founded in massachussetts not the west coast, which is why the original tweet says "formed".
What is the difference between "founded" and "formed" here?
To me, "Founded" feels like thats literally the initial point of creation. Formed feels more like a process.
To give an analogy, "Founded" would be like giving birth in Ohio. "Formed" would be like being raised in Kansas.
I didn't look up either definition though, that just what my instincts say
> if crime gets too bad, people simply don’t bother telling the police or other data-collecting bodies. The only crime that isn’t like this is murder.
So you're saying they don't tell data-collecting bodies but they do tell body-collecting data?
"the National Eye Institute did a study like this to prove that a $50 older drug worked just as well as a $2000 newer drug"
You're killing me, man; the word "prove" is simply not appropriate in this context. It's a bad habit of yours and it makes you look statistically naive. One of the first things I teach students in stats 101 is to not think in terms of proof but think in terms of strength of evidence. The distinction is not meaningless, and I am not an old man yelling at clouds. For example, in case you're not willing to take the words of some rando commenter:
https://statmodeling.stat.columbia.edu/2020/09/09/they-want-statistical-proof-whatever-that-is
Statistics is fundamentally about quantifying uncertainty, whereas proof exists in a domain with zero uncertainty (e.g. theorems). Again, this is *not* a trivial distinction and in fact is a rather large problem in science communication more generally.
That's not how the word is commonly understood.
In fact, the meaning you would ascribe to it is not even the most common one: to prove something in a court of law you can rely on evidence and testimony, for example, which I assume your definition would exclude.
I get being a stickler for specific definitions, but the word "proof/prove" is not exclusively (or even mostly) a technical term
I agree with TUM, no study can 'prove' equivalence of two drugs.
Who cares is that's how the word is "commonly" understood? Here it's being used in a specific context, namely a scientific context, in which my assertion is correct. No need for mental gymnastics to defend improper use of a word.
I'm sorry to have to push back on this, but I don't understand where you get your definition.
On the Oxford Handbook of Medical Statistics, for instance, there is no technical definition for the word "prove" or "proof". There are, however, usages of the word you would describe as incorrect. E.g 2.4, in the subsection "When consent may be withheld".:
“One solution in situations like these is for the researcher to decide in advance to offer the intervention to all control group subjects after the trial has finished, assuming that the intervention proves to be effective.”
In section 2.4, about placebos, we find the repeated expression “proven intervention”, in the phrase:
“The benefits, risks, burdens and effectiveness of a new intervention must be tested against those of the best current proven intervention”
How are all these things "proven" if studies cannot "prove" anything? Clearly the word is being used in its commonly understood meaning of: show to be true at a statistically significant threshold.
I get being a stickler for definitions, and maybe your approach would be better for communicating the difference, but that's not how it's being used in the technical literature right now, so I don't think you can be annoyed at people using it in this way.
EDIT: formatting and references
You argument appears to be "some other people are misusing the word too so it's not actually misused," and then you proceed to extrapolate that to "how it's being used in the technical literature right now."
I'm just going to facepalm and walk away for the sake of my blood pressure, and hope that perhaps more thoughtful people will take a second to consider their own limitations on the subject, and recognize that even very strong empirical evidence still leaves uncertainty and thereby precludes use of the word "proof" or any variant thereof. You know, the way actual statisticians do.
NCVS shouldn't have the reporting issues that FBI data does, and reports a similar reduction in property crimes.
E.g. https://www.researchgate.net/figure/NCVS-property-crime-rates-since-1973_fig1_228236639
> it would be too much of a coincidence for the (trauma-care-induced) decline in the murder rate to exactly correlate with the (recording-bias-induced) decline in the assault rate... Aaron argues that would-be murderers have adjusted by trying harder to kill their victims
"Wisdom of crowds" hypothesis amongst murderers...
#35: Women are typically said to have high rates of mental illness than men, especially teenage girls versus teenage boys. How much of this is the same reporting bias?
Given the political inclinations of teenage girls and boys, this might even be the same bias, not just the same type of bias!
On #35, the finding that liberals report their moods to be equal to conservatives doesn't hold up in other studies. I wrote about this here, where I also presented the results of an experiment investigating why there is an ideological mental health/mood gap in the first place. https://www.richardhanania.com/p/does-therapy-culture-explain-the
> The Caucuses [Wisconsia / Armechigan / Michibaijan]
This makes sense as a reinterpretation (into a form much more aligned with the general structure of English words) of an unknown name.
I always think it's funny that we call whites "Caucasian" in formal contexts, while...
(a) modern Americans tend to think, for no particular reason, that people from the region of the Caucasus shouldn't count as white;
(b) the Romans used "born from the Caucasus" as an extremely negative insult.
As I understand it, people consider Armenians white in most of the US *except* California, which has the largest Armenian population in the country. I doubt there are enough Georgians or Azeri in the US to get much attention in terms of racial classification.
When I was growing up in Connecticut there were three groups in town - Lebanese Christians, (mostly secular) Albanian Muslims, and Portuguese - who were “exotic whites,” meaning white in a racial sense but culturally distinct. In contrast, no one ever considered even the most light-skinned Puerto Ricans white.
(From the hydrofoil article: )
> A conventional boat pushes through water, and the resistance it encounters increases exponentially with speed. (Specifically: drag scales roughly with the cube of velocity, meaning that to double your speed you need eight times the power to overcome resistance.)
It's unusual to see a claim of exponential growth followed by an immediate proof that the growth is not exponential.
Well-spotted! That's an amusing misuse of it.
It’s not uncommon for people to misdescribe polynomials as exponential - after all, there’s an exponential there, and the curve is concave up!
There's an exponent; there isn't an exponential, which refers to a different type of object.
"Exponential growth" could be interpreted as referring to "growth in the exponent", with "exponential" as an adjective, or as "the rate of growth that characterizes exponentials", with "exponential" as a noun. (Both are defensible under the normal rules of English grammar, and the ambiguity can't really be resolved since those two definitions have identical meanings.)
Your explanation relies on the speaker not knowing any meaning at all for "exponential". This is often the case, but I don't see that your explanation can explain much. You wouldn't see someone confusing the noun "maniac" with the noun "mania", even though the derivational relationship is basically identical to the one between the noun "exponential" and the noun "exponent".
The explanation I'm thinking of is that people are interpreting "exponential" as meaning "having an exponent". They think that linear functions don't have an exponent (because we don't write down the 1), but think of both 2^x and x^2 as "exponential" functions because they have exponents.
After I stopped moaning in pain, I prompted Google search (and presumably Gemini, under the covers) with
>in the US, with normal schooling, at what IQ do half the population successfully learn what an exponential function is?
and it/they responded with:
>Based on IQ distribution and typical educational outcomes in the United States, half the population (roughly the 50th percentile) successfully learns and understands what an exponential function is (typically taught in Algebra II) at an IQ of approximately 100
( I _suspect_ that a higher IQ is actually required... - but maybe I'm cynical. )
Ask it when it went to school.
LOL! Cute! Many Thanks!
In regards to the Helen of Troy one, I am not sure that the Trojan War era Ancient Greeks actually understood how skin color and other "racial" characteristics were inherited. In stories of the Trojan War the Trojan ally, King Memnon of Ethiopia, is depicted as black in Ancient Greek artwork. However, he is also attested to be the son of Tithonus, a Trojan prince, and Eos, the goddess of the Dawn. Neither of these characters are depicted as black in ancient artwork.
The myths do say that Eos took Tithonus to live at "the ends of the Earth," which would imply somewhere in Africa since Memnon grew up to be an Ethiopian king. That implies that the Ancient Greeks at the time of Homer thought that anyone born in Africa would become black, even if they had white parents! So maybe Leda and Tyndareus were on a diplomatic mission to Africa when she had Helen (or rather, laid Helen's egg).
Fun fact: The Prose Edda identifies Memnon as the father of Thor. But also that Thor had 'hair fairer than gold' so presumably Snorri Sturluson wasn't completely familiar with the details of heredity either.
The darkness of Africans was mythologically explained by the sun god's son Pheaton's demand to control the sun for one day. He lost control and scorched part of the earth, creating the Sahara and black people. Perhaps Memnon was alive and in Ethiopia when this happened, and the Greeks believed in Lamarckism, so that his descendents would still inherit his dark skin.
Or they didn't think it through and just thought "every ethiopian king that I've hever heard of is black, if you tell me memnon is an Ethiopian king i'm going to depict him as such".
23. With respect to Lawyers and AI--my evaluation is that the use of AI in my practice (which is primarily regulatory and related to strategic counseling rather than litigation) is that AI is extremely useful, and the utility is not limited to ChatGPT 5.2 Pro or its ilk. A lot of firm work involves very quick turnaround and (for lack of a better word) dumbing down complex regulatory requirements into intuitive, gut level lists of "what you, the client, need to do."
After years of doing this, I usually know off the top of my head what the client needs to do with respect to the regulations that I work in, but I don't always have the exact citation and language off the top of my head. AI is excellent at fetching this from ecfr.gov, providing exact excerpts, and making this digestible for a client that is overwhelmed by regulatory complexity.
While the legal "know-how" that comes from intuition and general market and regulatory evaluation is, as the OP says, the real value that a lawyer provides, there is blackletter that is very much amenable to digestion via LLMs.
I am not a tax lawyer (though I work with them extensively), but I would imagine tax law would also be like this. Any realm of law that has to do with blackletter regulation and which can, therefore, probably be formalized will be the place where AI shines in the legal profession.
I once explained to Claude how I make good coffee in a basic Mr. Coffee. It was surprised it worked. It especially thought that using a microwave to heat the water to bloom the grounds was a bad idea.
Re: 28 (the LLM getting rewarded for using the calculator)
I hear that in some of its training runs, evolution rewarded people for looking at images of fertile young people of the opposite sex, and now lots of humans spend time looking at (or imagining) pictures of fertile young people of the opposite sex even while working on completely unrelated tasks!
Great now picking up my calculator makes me feel pervy.
What, you've ever typed 80085 on it before?
38, I read the linked blurb and this seems really confused. As “rights” talk, affirming or denying, tends to be IMO.
A “right”, practically speaking, is just what people and coalitions could make the wider culture they operate in accept as a “legitimate” want. “Power relations” etc.
E.g., people who can acquire some stuff through work and trade are generally more competent (and have been, e.g. in America, been very numerous), so getting to keep what you acquire that way is a “right” to property. Because many competent people would obviously just use violence to come to the same result otherwise, only much less pleasantly.
Inventing a new “right”, always a popular pastime, is just a claim to legitimacy of a want; if people (coalitions) can credibly make it stick, others will agree, otherwise, it’ll be met with something on the laughter…violence spectrum by the rest of society. Mostly, it stays with the middle ground of indirect, regulated, and institutional violence (voting).
Most people really, really want the culture they are used to to pervade their lives. Being a stranger in a strange land at home is disorienting and intensely painful to them. Outright majorities prefer marginal familiarity to marginal disposable wealth (if not, the real estate market would look vastly different).
Trying to keep moving policy in directions that very many, largely competent people deeply disagree with by calling their innermost wants illegitimate (“not rights”) is not likely to make people reconsider what they want, it’s just fighting words. So far the debate has escalated to “Trump”. It puzzles me that some people want to escalate it further; people will not want differently if scolded more.
It seems likely that improved trauma care could reduce the amount of (successful) murders by SOME amount.
It does NOT seem remotely likely that a 2x improvement in trauma care would consistently reduce murders _by a factor of 2_ (unless you quantify trauma care in a way that makes this tautological). The relationship between them seems obviously far less direct than that.
I wish people would put more thought into the actual quantitative relationship between the lever they are pulling and the outcome they are measuring.
> everyone in tech journalism was writing about Uber was
Presumably meant to read “writing about how Uber was” or “writing about Uber being”
Re 36 rooftops houses, we have some of these in San Francisco at Embarcadero Center. I even know someone who lived in one. They said it was reasonably nice.
However, WTF is up with the roads, cars, and presumably garages. Who is driving their car up there, how (10 minutes on ramps?), and why (the building doesn't have an express elevator for the luxury homes at the top?).
Victimisation rates are not subject to reporting bias, and they've been going down as well https://bjs.ojp.gov/library/publications/criminal-victimization-2024
My take on the lab leak change is that (1) it was a believable story that never had any solid evidence and (2) the X-is-destroying-America mob have moved on from China to maybe AI, Trump, America, or something else. The wind only blows in one direction for so long.
Does Scott use “cf” correctly?
Whenever Scott writes "cf" I read it as "c' ferinstance" (see, for instance), i.e. as a substitute for "e.g.", which I almost never see here.
Thank you. I’ll adopt the same translation and it will save me frequent momentary frustration. But put me down in favor of folks just using cite signals the normal way.
On 32 (the 60 minutes episode).
It wasn’t leaked. They forgot to stop it from airing on a streaming service in Canada…
#38: The word 'populism' has always confused me. Appealing to the mass of people with issues that they care about, in the hope of getting votes from them - this is just democracy working.
"But the mass of people doesn't even know what they want, they will vote for something that sounds good but turns out to be bad!"
Sure, and so says every out-of-touch, insulated and arrogant beaurocrat ever. History abounds with examples of people saying that the common people don't really want {x}, e.g. look at the speech "Slavery a Positive Good" (https://teachingamericanhistory.org/document/slavery-a-positive-good/) which argues that slavery is Good Actually and that the Mass of People who call for Abolition are Ignorant and Thoughtless and this will lead to the End of the Union.
Is the desire for regime change in Iran a 'populist' movement? If not, why not?
If so, is this good or bad?
Does this mean populism can both good or bad depending on what policies are being pushed (good populism, bad populism)?
If so, why is the word 'populism' overwhelmingly used negatively?
It's populism if it sounds good at first blush when you don't think about the consequences or have little knowledge on the topic. As most people don't spend much (any) time evaluating policies (very much separate from yelling about politics), the name fits.
The prototypical example is lowering taxes. Those who support it often didn't go over the financial plan of the country, they just see that their personal expenses will decrease on that one item and so they support it.
Obviously, when someone claims something is populism, they can be right or wrong, it's not a magic word. So someone claiming abolishing slavery is populism was simply wrong.
And you don't know whether you are right or wrong until much later.
So lowering taxes is populism, and bad.
However, if it turns out that lower tax rates leads to more economic activity and more wealth for all, then lowering taxes is retroactively good, and therefore it can't have been populism.
(The Laffer curve - some tax rates are just too high.)
It seems that calling a policy 'populist' is a slightly fancier way of saying 'I think it's bad'. You don't actually know the outcomes until they play out.
Sure you can predict and forecast and go off prior examples - but there is uncertainty in all of this, and that uncertainty and humility is reflected in the statement 'I think it's bad' but not in the statement 'that's populist so it's bad'.
It's more like "It's bad, but it seems good on low-informatiom.".
I don't see a problem with having a succint word for this and how much humility a speaker has seems nigh orthogonal to me and mostly depends on context. For example, one can also say "I think this policy is populist".
Even granting that, using populist for this is bad, because it causes confusion and makes the speaker look like an arrogant elitist who believes that only the highly credentialed can have meaningful opinions. When many "elite" opinions fit the definition way better. Including woke, DEI, and many other things.
> It's populism if it sounds good at first blush when you don't think about the consequences or have little knowledge on the topic.
This just sounds a bit like "It's populism if it's popular and I don't like it" though.
> So someone claiming abolishing slavery is populism was simply wrong
But we only know that with the benefit of hindsight, because we live in a world where slavery was (mostly) abolished and it worked out well.
Nah, I could have told you abolishing slavery was a good idea even before you Americans did it. Even if there had been severe economic implications, the inherent value of freeing people would have still outweighted it.
Prior to the Industrial Revolution, free societies tended to be conquered by extractive premodern states which enslaved or at least villeinized the majority of their population, and were able to raise professional armies on the base of that unfree labor.
The fact that liberal societies are viable today rests on our ability to outsource hard work to machines.
"It's populism if it sounds good at first blush when you don't think about the consequences or have little knowledge on the topic."
Oh? So if you do think about the consequences and decide that you're fine with them even if it makes your opponents extremely angry, it's no longer populism?
Take Brexit for an example: most people who voted for leaving the EU said they were fine with taking a moderate mid-to-long term economical hit if it meant clawing back control over their own laws and borders and ultimately throwing out all the brown immigrants. It didn't seem to stop any remain voters from calling Brexit populism, but you would argue they were wrong, then? Those people were simply seething losers?
I doubt the median voter actually conceptualized what moderate mid-to-long term economical hit and lower immegration entailed, but if they did then yes, it would have been wrong to call it populism.
I think part of the distinctiveness of popularism is that it is more explicitly anti-elite than just pro-popular stuff: it's not a focus on popular stuff in general, it's a focus on specifically the popular stuff that the elite don't like--sometimes maybe even conflating "not liked by the elite" with "popular in general" in a way that isn't justified.
Whether "populism" is reliably used to single out movements with this sort of inclination I'm not sure, but IMO that's a meaning of populism that I think distinguishes it from just democracy in general, and also one that makes it clear what the downsides could be: an inmate mistrust of elite institutions might mean a bad information ecosystem; mistrust of elites as a class means it's hard to accomplish things through regular political channels, since the will usually require collaboration or compromise with elites--meaning that there's a tendency to either diffuse but unproductive rage, or radical destabilizing change; etc.
Of course, a lot depends on what "elites" a movement is suspicious of: I'm more sympathetic to middle class lawyers and doctors being mad about kings and noblemen than I am to the "lumpenproletariat" of gamblers and mobsters being mad about judges and scientists, but of course YMMV.
"If so, why is the word 'populism' overwhelmingly used negatively?"
Because the professional middle class controls the means of communication and their views are at odds with those of the mass of the people on almost every point. Even their Fabian socialism is something they want to force onto the common man top-down for his own good. Chesterton wrote very lucidly about this already 120 years ago and nothing has changed. Orwell is also good on the problem of middle-class socialism, especially in Road to Wigan Pier.
It seems to me that you (and many people) are conflating a bit the meaning of "populism" with that of "demagoguery". The later means using a rhetoric that has a vast appeal using the prejudices of the mass instead of their rationality, and has a pejorative meaning per se.
Populism means using a rhetoric about (sometimes pretending) favoring the well-being of the "common people" to that of a (sometimes chimeric) elite. To me it is not derogatory per se, as there are indeed many society where there is an elite who favors its own interest going against that of the mass, for example indeed Iran. But in other context, this dichotomy between the mass and the elite is blurry, imaginary or cherry-picked (is the elite really the intellectual left rather than the billionaires?), or/and is applied to societies where such opposition wouldn't provide any benefit to the "common people".
Of course, demagoguery is often used in populistic rhetoric.
It's Ted Nasmith, not "Ted Naismith".
Facebook was famously founded in a Harvard dorm (and according to this - https://www.rollingstone.com/culture/culture-news/the-battle-for-facebook-242989/ - incorporated while still a dorm room operation).
Microsoft was founded in Albuquerque (see, e.g., "History of Microsoft" on Wikipedia). Gates and Allen dropped out of Harvard and moved to Albuquerque because the original plan was selling a BASIC interpreter to MITS, which was based in Albuquerque. The company was founded in 1976 and had 13 employees when it moved to Washington in 1979.
I judge the rest to be accurately West Coast, subject to the proviso that the Aramco thing is a bit murky among other reasons because Standard Oil of California was one of the successors of Standard Oil, which was founded in Ohio. Obviously Standard Oil's west coast acquisitions were originally founded on the west coast, but still...
Thanks for sharing my interview.
The recent furin cleavage site (FCS) finding in Brazil is of no importance. We already knew many betacoronaviruses have FCS. What we've never seen is a sarbecovirus FCS, and this evidence has only become stronger as more sarbecoviruses are discovered and still no FCS.
And of course, this rare FCS happens to appear as an insertion (very rare), and with the CGGCGG sequence that is nowhere else in the SARS2 genome.
"39: When complaining about modernity’s real and obvious flaws, it’s important not to forget how much lots of traditional societies sucked: “An Egyptian Muslim woman who lived under female seclusion since her marriage, 40 years ago, asks a female Christian missionary to describe flowers to her.”"
That's not a flaw of traditional societies. That's a flaw of fundamentalist Islam. Even under the Roman Empire, Christian societies never practiced female seclusion. In the 21st century, many Islamic societies still do.
If you ask chatGPT whether the medieval age was worse or better than the early modern age, for women, it produces an interesting synopsis
> I fact checked myself and Saramago's novel is not actually set in the middle ages, nor even the early modern period but, as it's the 18th century, it's the Age of Enlightenment.
Ah, although I suppose in terms of what you were talking about, nunneries, not much had changed.
> Granted Saramago's book is fiction, not History.
I'm not sure whether Saramago would consider nuns living in a nunnery as an example of societies practicing female seclusion. Would he have considered male monks as an example of male seclusion? If both monasteries and nunneries are considered as examples of seclusion in a society, can we say that society is characterized by female seclusion?
Also, more to the point, I believe that in Portugal - like in most medieval Christian nations - daughters were given a dowry at marriage. After receiving a dowry a daughter would have had no claim to an inheritance.
Not only are there flowers in convents, there are famously large quantities of them. Also, the type of women who in our day are vehemently feminist entered orders voluntarily in the middle ages, so it's not necessarily straightforward to distinguish between "seclusion" and "separatism".
"interpreting questions through a cultural lens rather than a Biblical one"
So, like actually existing religious communities do?
There's something very suspect about the World Happiness Report that I never got around to fully investigate. When you actually ask people directly whether they're happy, China wins by a margin [1]. Most notably, they're 83% positive on "country's social and political situation", compared to a world average of 40%. Interestingly, they stopped surveying China since.
This is a result that is fairly risky for a western institute to publish, and I always suspected that's why WHR engineer these weird questions that somehow bring Scandinavia on top. But I'm willing to be convinced otherwise.
Note that the common western view that "China threatens people to answer positively" is a silly myth not supported by evidence. Chinese have no problem telling pollsters that corrupt officials are the country's main problem - a much more sensitive issue than saying "I'm not happy". [2]
[1] https://www.ipsos.com/sites/default/files/ct/news/documents/2023-03/Ipsos%20Global%20Happiness%202023%20Report-WEB.pdf
[2] https://www.pewresearch.org/global/2015/09/24/corruption-pollution-inequality-are-top-concerns-in-china/
Do you have a source for China not being included in surveys? One of the authors is from the International Business School Suzhou, Xi'an Jiaotong-Liverpool University, which has a location in China. While I've only just skimmed the info, I see China mentioned in several chapters from the 2025 report. And looking at the data it seems like it's included.
https://www.worldhappiness.report/ed/2025/executive-summary/
Dashboard: https://data.worldhappiness.report/table
Or are you only referring to the direct question(s) you mention?
This was referring to the Ipsos poll [1] - China was no.1 in 2023, and since then they stopped polling there.
WHR do survey China, and the suggested hypothesis is they are deliberately removing questions that bring China on top.
Is Ipsos affiliated with the WHR mentioned in the OP links?
Skimming 2020-2025 WHR ranking data, I don't see China coming in the top 10 in any of the years. So it seems strange to me that they would juice their metrics to prevent something that's never happened in their previous reports.
As for Ipsos, I noticed that while China wasn't included in 2024 or 2025, though in 2024 at the bottom of some slides they state "The samples in Brazil, Chile, China, Colombia, Indonesia, Ireland, Malaysia, Mexico, Peru, Singapore, South Africa, Thailand, and Turkey are more urban, more educated, and/or more affluent than the general population."
That makes it seem like they had the China data, but didn't include it, or maybe just copy-pasted from a previous year's report.
From the downloadable report here: https://www.ipsos.com/en-us/global-happiness-2024
Don't take this the wrong way, but have you considered just asking them why they no longer include China? You could probably leverage Root Claim.
I'll ask myself, but I imagine you have much higher status than some internet rando.
Ipsos is a separate survey that asks the direct question. They're not related to WHR, who ask different questions that put China lower.
I didn't notice the comment you provided. I agree it's likely a copy-paste error in 2024.
Like I said, this is just some suspicion I have. I never really dug into it. In any case, Ipsos stopping surveying China is less important, since just the 2023 survey is enough to raise the suspicion that WHR (the recognized and respected happiness index everyone quotes) are hacking their results.
Re: the VeggieTails blog
It feels contrived to post a "Highlights from the comments..." followup post when there were only 34 comments to begin with.
re 19:
> I thought the commercials were in bad taste, misrepresenting what OpenAI’s ads would be like and turning the completely normal decision for a tech company to have an ad-supported free version of their product into some kind of horrible betrayal.
En contraire I thought it was a refreshingly honest statement (for a company) about the, pardon my lingo, surveillance capitalism that has been the primary business model for big tech of the last decades. Google betrays us by trapping the less tech-savvy (aka average) user into clicking on ad results. Meta betrays us by luring us with connections to our community while hooking us into content loops that leave us passive, overly-activated and eager to buy shit we don't need to give us relief.
I might be overstating and overdramatizing the case a bit, but I think the sentiment is growing and there is a lot of truth to be found in there. Add to that the alignment worries, with the misaligned over-optimizing for engagement machine that capitalism already is, even without viciously fast AI self-improvement iterations, and it looks like a very reasonable ad.
> I thought Sam Altman’s response was fair (although his countercriticism of Anthropic also missed the mark). People in his replies tried to enforce a norm of “if you write a long explanation defending yourself against someone else’s funny lies, that means you care and you lose”, but that’s a stupid norm and people should stop shoring it up
I only skimmed the comments on https://xcancel.com/sama/status/2019139174339928189, but they seemed more the usual level of X quippyness/brain-deadery to a tweet that read butthurt. At least from the view I got on it, I couldn't see any norm-enforcing (except for, don't be a bad loser)
Framing someone correcting implied falsehoods about their company's advertisement operations as being a bad loser is exactly the norm-enforcement Scott is talking about.
I gave arguments for why they are not falsehoods, his Alt’s post didn’t read as correcting anything to me, hence the only norm I’m enforcing is not being a bitter loser
20. Baseball pitchers throwing treason-style.
49. I assume the boat is now The True Self.
"DC has a rideshare app called Empower that charges 20-40% less than Uber. (Drivers like it too because they keep 100% of the fare)...DC is trying to shut it down because of liability insurance. DC law requires $1 million per ride. The $1 million requirement isn’t sized to typical accidents. When $100,000 is the limit available for an insurance claim, 96% of personal auto claims settle below $100,000...Empower can offer $7 rides partly because it circumvents the mandate. DC is shutting it down for exactly that reason.”
OK, but why doesn't Marvel get into the business?
8 - i am not a christian and have very few friends who are but the smug self-righteousness that non-christians point at christians bugs me. People who bend over backwards to avoid racial discrimination exhibit joyful religious discrimination. Who cares if the characters in a christian video can't be christians because they are not human. The vast majority of christians are just trying to be good people i believe - just leave them alone.
There has to be some tribalism in humans. Also, a lot of anti Christian sentiment is anti religious, but it’s not fashionable to criticise Islam.
I'd say that my experience is the opposite--most of the "anti-religious" sentiment I've seen was, when you looked closer, just hated of Christianity. Including most capital-A Atheist discourse. They had no problem with whatever half-understood pseudo eastern woo was currently fashionable. In fact, if you repackaged boring Christian thought into an eastern (or at least non obviously Christian) packaging, they'd swallow it whole.
> 8 - i am not a christian and have very few friends who are but the smug self-righteousness that non-christians point at christians bugs me.
Maybe you're just in the wrong circles, or noticing the wrong people? Within our own monkeysphere, you have the Psmiths, who are both Christian and write impossibly interesting, erudite, and articulate book reviews (as a couple, no less!), and which have been in Scott's blogroll and whose posts are commented on in open threads here.
When I worked in finance, one of my good friends was a Christian. Most of us were Phd's, most of us were data science or modeling or adjacent, and he was one of the smartest and most thoughtful among us, as well as a genuine role model in all he did. And as far as I could tell, everyone else respected him just as much as I did.
The book Bonds That Make Us Free, by C Terry Warner, is recommended by CFAR, and is a fairly religious book by an openly religious author.
I feel like I see numerous examples of people engaging thoughtfully and respectfully with Christians in our set.
Thanks for this. I did not mean to imply that everyone here has that smug attitude when it comes to christians. I did mean to imply that item 8, the one pointing out that a christian video series, VeggiTales, is stupid because the characters are not in fact christians, seems to me to exhibit the small-minded tribal prejudice against christians that i see in our society. I don't like it.
Did we read the same blog post? I parsed it as a respectful, nerd-enthusiastic account of how the show's Christian creator had in fact been remarkably consistent in his application of his own dogma, contrary to the claims of his (less insightful, also Christian) detractors who were calling the show blasphemy.
RE: 14
I'm afraid it took me too long to read the link and now most people will have read it, but I would really like to push back on this. The linked article is Not Very Good, and the article it's responding to is at least a bit more interesting.
I'm saying it's "Not Very Good" because it doesn't address most points raised in the original, and it has an annoyingly dismissive tone; like when it accuses the original article of being "a little rhetorical" and that the audience "responded to trigger words like 'elite misinformation'".
The first one is annoying in its own right, since the tone of blogs is often halfway between the informative and the personal; I'm not surprised to see references to personal opinions or perceptions of a matter, especially when they serve as hooks to discuss the actual analysis.
The second one is annoyingly taken out of context, and reading where it appears in the original gives a much different impression of the piece. The quoted section appears in the concluding paragraph:
“Over the last years, media outlets like the New York Times, universities like Oxford, and international institutions like the UN have devoted themselves to the fight against so-called “misinformation.” It is certainly true that our political discourse is awash with dangerous distortions and outright lies. But any institution which wishes to address that problem must start by looking into the mirror—and cease spreading “elite misinformation” like the World Happiness Report.”
The author is drawing a parallel between the scaremongering about "misinformation" done by traditional information institutions and their own irresponsible spreading of half-truths, which has been proven (according to the author at least) in the rest of the piece.
It's needlessly inflammatory to simply quote the part about "elite misinformation", which elicits the image of a conspiratorial mind, when the author is actually accusing traditional outlets of writing clickbait, which is a much more reasonable accusation.
Yascha doesn’t accuse traditional outlets of writing “clickbait;” he accuses them of spreading “misinformation,” which he seems to equate with “dangerous distortions and outright lies.” This is known as “poisoning the well;” by declaring the World Happiness report to be misinformation and a sham, he is preemptively declaring that anyone who claims to disagree with him is lying. Huw has one paragraph where he (charitably, in my view) says that Yascha’s post is “a little rhetorical,” but he otherwise attempts to address the substance of Yascha’s and ignores the rhetoric.
I suspect that Scott linked to the article in part because it was a high effort response, by which I mean that Huw identified Yascha’s major arguments in support of the thesis that “the world happiness report is a sham” and responded to all of them. You write that Huw, “doesn't address most points raised in the original,” but rather than list the points yuou think that Huw failed to address, you devote all of your seven paragraphs to complaining about the tone of a single paragraph by Huw. So we will have to agree to disagree on this.
That's fair, I didn't get to the substance because I was put off by the tone of the reply. But I would still strongly argue that the posted reply doesn't do justice to the original piece at all.
For instance: the original points out that different researchers publishing happiness scores can arrive at very different scores for the countries involved, despite ostensibly measuring happiness. If you had two IQ tests that gave out very different results, you would rightly be skeptical of the efficacy of the measurement.
I guess in part I am mixing in my own opinions about the topic, because Yascha's article doesn't spell this out in so many words, but it's clear from his bringing up a different happiness score that gives different results as a counterpoint to the validity of the World Happiness.
Also, this is in the reply article:
It is not clear to me that finding a bias in the ladder contradicts its purpose. Is it not reasonable that having money and being successful might make people satisfied with their lives? (Again, the individual-level results in Tables 11–13 find that life satisfaction correlates with income, and also age, gender, freedom, marital status, social support, health, institutional trust, etc.).
This is clearly circular reasoning: bias in the measurement is not a big deal because... the measurement is found to correlate with the items it is biased in favor of?? Obviously if you skew the questions to give more weight to economical factors it will correlate with those; same for the other biases.
I reread the original article and I concede that the tone can be annoying, especially if you disagree with the premises, but the answer isn't really much better, complaining about the tone of the original while making logical mistakes and just shrugging at the valid concerns.
(concerns much better expressed in that one _sam[ ]zdat_ post)
I went back again to the original article, but I'm starting to suspect we are reading two different pages. The one I read has the key takeaway in the last two paragraphs that read:
More broadly, supposedly serious news outlets still have a long way to go in subjecting publicity exercises like the World Happiness Report to appropriate journalistic scrutiny. It is easy to see why ***editors are tempted to assign some beat reporter without expertise in the social sciences to write up a fun little story about how much happier those enlightened Scandinavians are compared to benighted Americans.*** But if the media wants to live up to its self-appointed role as a gatekeeper of reliable information, it can’t continue to be complicit in the spread of such ***shoddy clickbait.***
Over the last years, media outlets like the New York Times, universities like Oxford, and international institutions like the UN have devoted themselves to the fight against so-called “misinformation.” It is certainly true that our political discourse is awash with dangerous distortions and outright lies. But any institution which wishes to address that problem must start by looking into the mirror—and cease spreading “elite misinformation” like the World Happiness Report.
(apologies for the lack of formatting, I don't know how to highlight stuff)
Again, it is clear from the phrasing that the accusation is that journalists are tempted to write according to simple narratives rather than research a topic accurately; this lack of research shows up as shoddy clickbait; again, the point made in the piece is that the outlets that pick up the news are latching onto the idea that the US is falling behind on "happiness", and should learn from northern Europe. It's not hard to see the motivated reasoning behind this choice.
We don't really have to agree to disagree on the specifics though: the response post addresses none of this, it just defends the measurement as flawed but useful, which is tangential at best to the original post. It is high-ish effort, but it doesn't identify the major arguments nor does it address all of them (at least not satisfactorily; see the logical contradiction in hand-waving away the reported bias in Cantril ladder measurements)
My mistake; I reread the last paragraph before posting but didn’t check the preceding which talks about clickbait.
Huw is writing about the thesis: “The World Happiness Report Is a Sham.” That’s the title of Yascha’s piece, so Huw is not nitpicking by focusing on it. He ignores the media criticism except to say that it doesn’t support that theseis: “The WHR are extremely clear throughout their report about their methodology, and present explanations of it in a long preamble before presenting their annual table; it is simply not their fault if media outlets looking for a story obscure this.” So, yes, there is a bunch of stuff that Huw does not address.
With regard to Yasha’s contention that, “the Cantril Ladder...does not even do a good job of measuring respondents’ satisfaction with their own lives,” his support for this is research that concludes that, “the Cantril Ladder’s structure appears to influence participants to attend to a more power and wealth oriented view of well-being.” Two problems. First, “appears to influence” doesn’t support a definitive assertion; at best it would support “appears not to do a good job.”
Second, the implicit claim being made here is that, “a more power and wealth oriented view of well-being” is a poor measure of well-being. Huw disagrees. His position is that, “self-reported wellbeing measures are accurate by definition,” which he follows with a bunch of qualifications. I don’t find that completely convincing because it’s defining the problem away. Huw allows that results thrown off by an anchoring effect of previous questions. It’s not clear why that is different, in principle, from the wording of the question itself throwing off the results.
However, Yasha made the implicit claim, so the burden of proof is on him. He provides no argument, convincing or otherwise, to support it.
I'll update my belief that the original is less-than-fine, but I still think the answer is a bit lacking, although you raised some good points about the prominence of the sham-conspiratorial angle.
If you're interested in reading more about the topic, though, I have to recomment the sam[]zdat article I mentioned before. I think you might find the style extremely grating, so it's possible you'll deeply dislike it, but I find the author's position to be strongly argued and well supported. Cheers!
https://samzdat.com/2018/08/22/love-and-happiness/
#49: I wonder if this story is quite right. The initial SF Chronicle piece that broke the Ellison yacht story at least implied that the name came from the previous owner, which Ellison kept until realizing the name issue: "1999-11-05 04:00:00 PDT NORTH BAY -- Sailors who hang around the Sausalito harbor were admiring Larry Ellison's newest yacht, the $25 million "Izanami" docked at Schoonmaker Marina. Noticing that the 192-foot boat was made in Germany, they were somewhat shocked when they figured out what "Izanami" is backwards.
But we learned that the previous owner is Japanese and that "Izanami" is a goddess in Japanese mythology. Anyway, Larry, if you want a new name, we suggest "El Caro," the expensive one. It works forwards and backwards."
The brief mention in "Softwar" does not clear this up
I wonder if historical boat naming records are available to confirm or disconfirm this
https://www.sfchronicle.com/bayarea/article/Oracle-Chief-s-Yacht-Harbors-A-Surprise-2898468.php
"I’ll take this opportunity to pitch my startup idea - a dating site where, instead of checking boxes to see if you match, you give a willingness-to-date between 0 and 9, and match if your combined WTD is 10 or greater (so it could be both people rating the other 5, or you rating them 9 and them rating you 1, and so on)."
I like this idea but worry it would fall victim to "rateflation" in the same way, e.g, customer service ratings do, where 5 stars is the default. I understand the incentives here are different since votes are (essentially) private (although if you rate someone a 1 and you match, you know they rated you a 9), but I still think most people are just uncomfortable with eating people they like low, or the general idea of putting how much they like someone into a number. Not to mention if you just rate everyone a 10 you get more matches, which incentives desperate people in the same way other systems do. Also generally people probably want to match with people who like them about as much as they like them, not huge asymmetries. Probably a workable concept with some iteration, though!